Automotive DDR4 SDRAM: MT40A512M8 MT40A256M16 Features

4Gb: x8, x16 Automotive DDR4 SDRAM
Features
Automotive DDR4 SDRAM

MT40A512M8
MT40A256M16
Features • AEC-Q100
• PPAP submission
• VDD = V DDQ = 1.2V ±60mV
• VPP = 2.5V –125mV/+250mV Options1 Marking
• On-die, internal, adjustable V REFDQ generation • Configuration
• 1.2V pseudo open-drain I/O – 512 Meg x 8 512M8
• Refresh maximum interval time at T C temperature – 256 Meg x 16 256M16
range: • BGA package (Pb-free) – x8
– 64ms at –40°C to 85°C – 78-ball (9mm x 10.5mm) – Rev. B RH
– 32ms at 85°C to 95°C – 78-ball (7.5mm x 11mm) – Rev. F SA
– 16ms at 96°C to 105°C • FBGA package (Pb-free) – x16
– 8ms at 106°C to 125°C – 96-ball (9mm x 14mm) – Rev. B GE
• 16 internal banks ( x8): 4 groups of 4 banks each – 96-ball (7.5mm x 13.5mm) – Rev. F LY
• 8 internal banks (x16): 2 groups of 4 banks each • Timing – cycle time
• 8n-bit prefetch architecture – 0.625ns @ CL = 22 (DDR4-3200) -062E
• Programmable data strobe preambles – 0.750ns @ CL = 18 (DDR4-2666) -075E
• Data strobe preamble training – 0.833ns @ CL = 16 (DDR4-2400) -083E
• Command/Address latency (CAL) • Product certification
• Multipurpose register read and write capability – Automotive A
• Write leveling • Operating temperature
• Self refresh mode – Industrial (–40°C ื T C ื +95°C) IT
• Low-power auto self refresh (LPASR) – Automotive (–40°C ื T C ื +105°C) AT
• Temperature controlled refresh (TCR) – Ultra-high (–40°C ื T C ื +125°C)3 UT
• Fine granularity refresh – Revision :B :F
• Self refresh abort Notes: 1. Not all options listed can be combined to
• Maximum power saving define an offered product. Use the part cat-
• Output driver calibration alog search on http://www.micron.com for
• Nominal, park, and dynamic on-die termination available offerings.
(ODT) 2. The ×4 device is not offered and the mode
• Data bus inversion (DBI) for data bus is not supported by the x8 or x16 device
• Command/Address (CA) parity even though some ×4 mode descriptions ex-
• Databus write cyclic redundancy check (CRC) ist in the datasheet.
• Per-DRAM addressability 3. The UT option use based on automotive us-
• Connectivity test age model. Please contact Micron sales rep-
• Hard post package repair (hPPR) and soft post resentative if you have questions.
package repair (sPPR) modes 4. -062E is only available for die Rev. F.
• JEDEC JESD-79-4 compliant
CCMTD-1725822587-10418 Micron Technology, Inc. reserves the right to change products or specifications without notice.
4gb_auto_ddr4_sdram_z90b_z10B.pdf - Rev. I 05/19 EN 1 2016 Micron Technology, Inc. All rights reserved.
Products and specifications discussed herein are subject to change by Micron without notice.
Features
Table 1: Key Timing Parameters
Speed Grade Data Rate (MT/s) Target CL-nRCD-nRP tAA (ns) tRCD (ns) tRP (ns)
-062E1 3200 22-22-22 13.75 13.75 13.75
-075E1 2666 18-18-18 13.5 13.5 13.5
-083E 2400 16-16-16 13.32 13.32 13.32
Note: 1. Refer to the Speed Bin Tables for backward compatibility
Table 2: Addressing
Parameter 512 Meg x 8 256 Meg x 16

Number of bank groups 4 2
Bank group address BG[1:0] BG0
Bank count per group 4 4
Bank address in bank group BA[1:0] BA[1:0]
Row addressing 32K (A[14:0]) 32K (A[14:0])
Column addressing 1K (A[9:0]) 1K (A[9:0])
Page size1 1KB 2KB
Note: 1. Page size is per bank, calculated as follows:

Page size = 2COLBITS × ORG/8, where COLBIT = the number of column address bits and ORG = the number of
DQ bits.
Features
Figure 1: Order Part Number Example
([DPSOH3DUW1XPEHU 07$05+($$7%

07$ &RQILJXUDWLRQ 3DFNDJH 6SHHG 5HYLVLRQ
^
&RQILJXUDWLRQ 0DUN 5HYLVLRQ
0HJ[ 0 %)
0HJ[ 0
&DVH7HPSHUDWXUH 0DUN
3DFNDJH 0DUN &RPPHUFLDO 1RQH
EDOOPP[PP)%*$ 5+ ,QGXVWULDOWHPSHUDWXUH ,7
EDOOPP[PP)%*$ 6$ $XWRPRWLYH $7
EDOOPP[PP)%*$ *( 8OWUDKLJK 87
EDOOPP[PP)%*$ /<
3URGXFWFHUWLILFDWLRQ 0DUN
$XWRPRWLYH $
0DUN 6SHHG*UDGH
( W&. QV&/
( W&. QV&/

( W&. QV&/
Features
Contents
Important Notes and Warnings ....................................................................................................................... 19
General Notes and Description ....................................................................................................................... 19
Description ................................................................................................................................................ 19
Industrial Temperature ............................................................................................................................... 20
Automotive Temperature ............................................................................................................................ 20
Ultra-high Temperature .............................................................................................................................. 20
General Notes ............................................................................................................................................ 20
Definitions of the Device-Pin Signal Level ................................................................................................... 21
Definitions of the Bus Signal Level ............................................................................................................... 21
Functional Block Diagrams ............................................................................................................................. 22
Ball Assignments ............................................................................................................................................ 23
Ball Descriptions ............................................................................................................................................ 25
Package Dimensions ....................................................................................................................................... 28
State Diagram ................................................................................................................................................ 32
Functional Description ................................................................................................................................... 34
RESET and Initialization Procedure ................................................................................................................. 35
Power-Up and Initialization Sequence ......................................................................................................... 35
RESET Initialization with Stable Power Sequence ......................................................................................... 38
Uncontrolled Power-Down Sequence .......................................................................................................... 39
Programming Mode Registers ......................................................................................................................... 40
Mode Register 0 .............................................................................................................................................. 43
Burst Length, Type, and Order ..................................................................................................................... 44
CAS Latency ............................................................................................................................................... 45
Test Mode .................................................................................................................................................. 46
Write Recovery (WR)/READ-to-PRECHARGE ............................................................................................... 46
DLL RESET ................................................................................................................................................. 46
Mode Register 1 .............................................................................................................................................. 47
DLL Enable/DLL Disable ............................................................................................................................ 48
Output Driver Impedance Control ............................................................................................................... 49
ODT RTT(NOM) Values .................................................................................................................................. 49
Additive Latency ......................................................................................................................................... 49
Rx CTLE Control ......................................................................................................................................... 49
Write Leveling ............................................................................................................................................ 50
Output Disable ........................................................................................................................................... 50
Termination Data Strobe ............................................................................................................................. 50
Mode Register 2 .............................................................................................................................................. 51
CAS WRITE Latency .................................................................................................................................... 53
Low-Power Auto Self Refresh ....................................................................................................................... 53
Dynamic ODT ............................................................................................................................................ 53
Write Cyclic Redundancy Check Data Bus .................................................................................................... 53
Mode Register 3 .............................................................................................................................................. 54
Multipurpose Register ................................................................................................................................ 55
WRITE Command Latency When CRC/DM is Enabled ................................................................................. 56
Fine Granularity Refresh Mode .................................................................................................................... 56
Temperature Sensor Status ......................................................................................................................... 56
Per-DRAM Addressability ........................................................................................................................... 56
Gear-Down Mode ....................................................................................................................................... 56
Mode Register 4 .............................................................................................................................................. 57
Hard Post Package Repair Mode .................................................................................................................. 58
Soft Post Package Repair Mode .................................................................................................................... 58
Features
WRITE Preamble ........................................................................................................................................ 59

READ Preamble .......................................................................................................................................... 59
READ Preamble Training ............................................................................................................................ 59
Temperature-Controlled Refresh ................................................................................................................. 59
Command Address Latency ........................................................................................................................ 59
Internal V REF Monitor ................................................................................................................................. 59
Maximum Power Savings Mode ................................................................................................................... 60
Mode Register 5 .............................................................................................................................................. 61
Data Bus Inversion ..................................................................................................................................... 62
Data Mask .................................................................................................................................................. 63
CA Parity Persistent Error Mode .................................................................................................................. 63
ODT Input Buffer for Power-Down .............................................................................................................. 63
CA Parity Error Status ................................................................................................................................. 63
CRC Error Status ......................................................................................................................................... 63
CA Parity Latency Mode .............................................................................................................................. 63
Mode Register 6 .............................................................................................................................................. 64
tCCD_L Programming ................................................................................................................................. 65
VREFDQ Calibration Enable .......................................................................................................................... 65

VREFDQ Calibration Range ........................................................................................................................... 65
VREFDQ Calibration Value ............................................................................................................................ 66
Truth Tables ................................................................................................................................................... 67
NOP Command .............................................................................................................................................. 70
DESELECT Command .................................................................................................................................... 70
DLL-Off Mode ................................................................................................................................................ 70
DLL-On/Off Switching Procedures .................................................................................................................. 72
DLL Switch Sequence from DLL-On to DLL-Off ........................................................................................... 72
DLL-Off to DLL-On Procedure .................................................................................................................... 74
Input Clock Frequency Change ....................................................................................................................... 75
Write Leveling ................................................................................................................................................ 76
DRAM Setting for Write Leveling and DRAM TERMINATION Function in that Mode ..................................... 77
Procedure Description ................................................................................................................................ 78
Write Leveling Mode Exit ............................................................................................................................ 79
Command Address Latency ............................................................................................................................ 81
Low-Power Auto Self Refresh Mode ................................................................................................................. 86
Manual Self Refresh Mode .......................................................................................................................... 86
Multipurpose Register .................................................................................................................................... 88
MPR Reads ................................................................................................................................................. 89
MPR Readout Format ................................................................................................................................. 91
MPR Readout Serial Format ........................................................................................................................ 91
MPR Readout Parallel Format ..................................................................................................................... 92
MPR Readout Staggered Format .................................................................................................................. 93
MPR READ Waveforms ............................................................................................................................... 94
MPR Writes ................................................................................................................................................ 96
MPR WRITE Waveforms .............................................................................................................................. 97
MPR REFRESH Waveforms ......................................................................................................................... 98
Gear-Down Mode .......................................................................................................................................... 101
Maximum Power-Saving Mode ....................................................................................................................... 104
Maximum Power-Saving Mode Entry .......................................................................................................... 104
Maximum Power-Saving Mode Entry in PDA .............................................................................................. 105
CKE Transition During Maximum Power-Saving Mode ................................................................................ 105
Maximum Power-Saving Mode Exit ............................................................................................................ 105
Command/Address Parity .............................................................................................................................. 107
Features
Per-DRAM Addressability .............................................................................................................................. 115

VREFDQ Calibration ........................................................................................................................................ 118
VREFDQ Range and Levels ........................................................................................................................... 119
VREFDQ Step Size ........................................................................................................................................ 119
VREFDQ Increment and Decrement Timing .................................................................................................. 120
VREFDQ Target Settings ............................................................................................................................... 124
Connectivity Test Mode ................................................................................................................................. 126
Pin Mapping ............................................................................................................................................. 126
Minimum Terms Definition for Logic Equations ......................................................................................... 127
Logic Equations for a ×4 Device .................................................................................................................. 127
Logic Equations for a ×8 Device .................................................................................................................. 128
Logic Equations for a ×16 Device ................................................................................................................ 128
CT Input Timing Requirements .................................................................................................................. 128
Excessive Row Activation ............................................................................................................................... 130
Post Package Repair ....................................................................................................................................... 131
Post Package Repair ................................................................................................................................... 131
Hard Post Package Repair .......................................................................................................................... 132
hPPR Row Repair - Entry ........................................................................................................................ 132
hPPR Row Repair – WRA Initiated (REF Commands Allowed) .................................................................. 132
hPPR Row Repair – WR Initiated (REF Commands NOT Allowed) ............................................................. 134
sPPR Row Repair ....................................................................................................................................... 135
hPPR/sPPR Support Identifier .................................................................................................................... 138
ACTIVATE Command .................................................................................................................................... 139
PRECHARGE Command ................................................................................................................................ 140
REFRESH Command ..................................................................................................................................... 140
Temperature-Controlled Refresh Mode .......................................................................................................... 143
Normal Temperature Mode ........................................................................................................................ 143
Extended Temperature Mode ..................................................................................................................... 143
Fine Granularity Refresh Mode ....................................................................................................................... 146
Mode Register and Command Truth Table .................................................................................................. 146
tREFI and tRFC Parameters ........................................................................................................................ 146
Changing Refresh Rate ............................................................................................................................... 149

Usage with TCR Mode ................................................................................................................................ 149
Self Refresh Entry and Exit ......................................................................................................................... 149
SELF REFRESH Operation .............................................................................................................................. 151
Self Refresh Abort ...................................................................................................................................... 153
Self Refresh Exit with NOP Command ......................................................................................................... 154
Power-Down Mode ........................................................................................................................................ 156
Power-Down Clarifications – Case 1 ........................................................................................................... 161
Power-Down Entry, Exit Timing with CAL ................................................................................................... 162
ODT Input Buffer Disable Mode for Power-Down ............................................................................................ 164
CRC Write Data Feature ................................................................................................................................. 166
CRC Write Data ......................................................................................................................................... 166
WRITE CRC DATA Operation ...................................................................................................................... 166
DBI_n and CRC Both Enabled .................................................................................................................... 167
DM_n and CRC Both Enabled .................................................................................................................... 167
DM_n and DBI_n Conflict During Writes with CRC Enabled ........................................................................ 167
CRC and Write Preamble Restrictions ......................................................................................................... 167
CRC Simultaneous Operation Restrictions .................................................................................................. 167
CRC Polynomial ........................................................................................................................................ 167
CRC Combinatorial Logic Equations .......................................................................................................... 168
Burst Ordering for BL8 ............................................................................................................................... 169
Features
CRC Data Bit Mapping ............................................................................................................................... 169

CRC Enabled With BC4 .............................................................................................................................. 170
CRC with BC4 Data Bit Mapping ................................................................................................................ 170
CRC Equations for x8 Device in BC4 Mode with A2 = 0 and A2 = 1 ................................................................ 173
CRC Error Handling ................................................................................................................................... 174
CRC Write Data Flow Diagram ................................................................................................................... 176
Data Bus Inversion ........................................................................................................................................ 177
DBI During a WRITE Operation .................................................................................................................. 177
DBI During a READ Operation ................................................................................................................... 178
Data Mask ..................................................................................................................................................... 179
Programmable Preamble Modes and DQS Postambles .................................................................................... 181
WRITE Preamble Mode .............................................................................................................................. 181
READ Preamble Mode ............................................................................................................................... 184
READ Preamble Training ........................................................................................................................... 184
WRITE Postamble ...................................................................................................................................... 185
READ Postamble ....................................................................................................................................... 185
Bank Access Operation .................................................................................................................................. 187
READ Operation ............................................................................................................................................ 191
Read Timing Definitions ............................................................................................................................ 191
Read Timing – Clock-to-Data Strobe Relationship ....................................................................................... 192
Read Timing – Data Strobe-to-Data Relationship ........................................................................................ 194
tLZ(DQS), tLZ(DQ), tHZ(DQS), and tHZ(DQ) Calculations ............................................................................ 195
tRPRE Calculation ..................................................................................................................................... 196
tRPST Calculation ...................................................................................................................................... 197
READ Burst Operation ............................................................................................................................... 198

READ Operation Followed by Another READ Operation .............................................................................. 200
READ Operation Followed by WRITE Operation .......................................................................................... 205
READ Operation Followed by PRECHARGE Operation ................................................................................ 211
READ Operation with Read Data Bus Inversion (DBI) .................................................................................. 214
READ Operation with Command/Address Parity (CA Parity) ........................................................................ 215
READ Followed by WRITE with CRC Enabled .............................................................................................. 217
READ Operation with Command/Address Latency (CAL) Enabled ............................................................... 218
WRITE Operation .......................................................................................................................................... 220
Write Timing Definitions ........................................................................................................................... 220
Write Timing – Clock-to-Data Strobe Relationship ...................................................................................... 220
tWPRE Calculation .................................................................................................................................... 222
tWPST Calculation ..................................................................................................................................... 223
Write Timing – Data Strobe-to-Data Relationship ........................................................................................ 223

WRITE Burst Operation ............................................................................................................................. 227
WRITE Operation Followed by Another WRITE Operation ........................................................................... 229
WRITE Operation Followed by READ Operation .......................................................................................... 235
WRITE Operation Followed by PRECHARGE Operation ............................................................................... 239
WRITE Operation with WRITE DBI Enabled ................................................................................................ 242
WRITE Operation with CA Parity Enabled ................................................................................................... 244
WRITE Operation with Write CRC Enabled ................................................................................................. 245
Write Timing Violations ................................................................................................................................. 250
Motivation ................................................................................................................................................ 250
Data Setup and Hold Violations ................................................................................................................. 250
Strobe-to-Strobe and Strobe-to-Clock Violations ........................................................................................ 250
ZQ CALIBRATION Commands ....................................................................................................................... 251
On-Die Termination ...................................................................................................................................... 253
ODT Mode Register and ODT State Table ........................................................................................................ 253
Features
ODT Read Disable State Table .................................................................................................................... 254

Synchronous ODT Mode ................................................................................................................................ 255
ODT Latency and Posted ODT .................................................................................................................... 255
Timing Parameters .................................................................................................................................... 255
ODT During Reads .................................................................................................................................... 257
Dynamic ODT ............................................................................................................................................... 258
Functional Description .............................................................................................................................. 258
Asynchronous ODT Mode .............................................................................................................................. 261
Electrical Specifications ................................................................................................................................. 262
Absolute Ratings ........................................................................................................................................ 262
DRAM Component Operating Temperature Range ...................................................................................... 262
Electrical Characteristics – AC and DC Operating Conditions .......................................................................... 263
Supply Operating Conditions ..................................................................................................................... 263
Leakages ................................................................................................................................................... 264
VREFCA Supply ............................................................................................................................................ 264
VREFDQ Supply and Calibration Ranges ....................................................................................................... 265
VREFDQ Ranges ........................................................................................................................................... 266
Electrical Characteristics – AC and DC Single-Ended Input Measurement Levels .............................................. 267
RESET_n Input Levels ................................................................................................................................ 267
Command/Address Input Levels ................................................................................................................ 267
Command, Control, and Address Setup, Hold, and Derating ........................................................................ 269
Data Receiver Input Requirements ............................................................................................................. 271
Connectivity Test (CT) Mode Input Levels .................................................................................................. 275
Electrical Characteristics – AC and DC Differential Input Measurement Levels ................................................. 279
Differential Inputs ..................................................................................................................................... 279
Single-Ended Requirements for CK Differential Signals ............................................................................... 280
Slew Rate Definitions for CK Differential Input Signals ................................................................................ 281
CK Differential Input Cross Point Voltage .................................................................................................... 282
DQS Differential Input Signal Definition and Swing Requirements .............................................................. 283
DQS Differential Input Cross Point Voltage ................................................................................................. 285
Slew Rate Definitions for DQS Differential Input Signals .............................................................................. 286
Electrical Characteristics – Overshoot and Undershoot Specifications ............................................................. 288
Address, Command, and Control Overshoot and Undershoot Specifications ................................................ 288
Clock Overshoot and Undershoot Specifications ......................................................................................... 289
Data, Strobe, and Mask Overshoot and Undershoot Specifications .............................................................. 290
Electrical Characteristics – AC and DC Output Measurement Levels ................................................................ 290
Single-Ended Outputs ............................................................................................................................... 290
Differential Outputs .................................................................................................................................. 292
Reference Load for AC Timing and Output Slew Rate ................................................................................... 293
Connectivity Test Mode Output Levels ........................................................................................................ 294
Electrical Characteristics – AC and DC Output Driver Characteristics ............................................................... 295
Connectivity Test Mode Output Driver Electrical Characteristics ................................................................. 295
Output Driver Electrical Characteristics ..................................................................................................... 297
Output Driver Temperature and Voltage Sensitivity ..................................................................................... 300
Alert Driver ............................................................................................................................................... 300
Electrical Characteristics – On-Die Termination Characteristics ...................................................................... 301
ODT Levels and I-V Characteristics ............................................................................................................ 301
ODT Temperature and Voltage Sensitivity ................................................................................................... 303
ODT Timing DefinitionsODT Timing Definitions and Waveforms ................................................................ 303
DRAM Package Electrical Specifications ......................................................................................................... 307
Thermal Characteristics ................................................................................................................................. 311
Current Specifications – Measurement Conditions .......................................................................................... 312
Features
IDD, IPP, and IDDQ Measurement Conditions ................................................................................................ 312

IDD Definitions .......................................................................................................................................... 314
Current Specifications – Patterns and Test Conditions ..................................................................................... 318
Current Test Definitions and Patterns ......................................................................................................... 318
IDD Specifications ...................................................................................................................................... 327
Current Specifications – Limits ....................................................................................................................... 328
Speed Bin Tables ........................................................................................................................................... 338
Backward Compatibility ............................................................................................................................ 338
Refresh Parameters By Device Density ............................................................................................................ 357
Electrical Characteristics and AC Timing Parameters – DDR4-1600 through DDR4-2400 ................................... 358
Electrical Characteristics and AC Timing Parameters – DDR4-2666 through DDR4-3200 ................................... 370
Converting Time-Based Specifications to Clock-Based Requirements .............................................................. 382
Options Tables .............................................................................................................................................. 383
Features
List of Figures
Figure 1: Order Part Number Example .............................................................................................................. 3
Figure 2: 512 Meg x 8 Functional Block Diagram ............................................................................................. 22
Figure 3: 256 Meg x 16 Functional Block Diagram ........................................................................................... 22
Figure 4: 78-Ball x4, x8 Ball Assignments ........................................................................................................ 23
Figure 5: 96-Ball x16 Ball Assignments ............................................................................................................ 24
Figure 6: 78-Ball FBGA – ×4, ×8, "RH" ............................................................................................................. 28
Figure 7: 78-Ball FBGA – ×4, ×8, "SA" .............................................................................................................. 29
Figure 8: 96-Ball FBGA – ×16, "GE" ................................................................................................................. 30
Figure 9: 96-Ball FBGA – ×16, "LY" .................................................................................................................. 31
Figure 10: Simplified State Diagram ............................................................................................................... 32
Figure 11: RESET and Initialization Sequence at Power-On Ramping ............................................................... 38
Figure 12: RESET Procedure at Power Stable Condition ................................................................................... 39
Figure 13: tMRD Timing ................................................................................................................................ 41
Figure 14: tMOD Timing ................................................................................................................................ 41
Figure 15: DLL-Off Mode Read Timing Operation ........................................................................................... 71
Figure 16: DLL Switch Sequence from DLL-On to DLL-Off .............................................................................. 73
Figure 17: DLL Switch Sequence from DLL-Off to DLL-On .............................................................................. 74
Figure 18: Write Leveling Concept, Example 1 ................................................................................................ 76
Figure 19: Write Leveling Concept, Example 2 ................................................................................................ 77
Figure 20: Write Leveling Sequence (DQS Capturing CK LOW at T1 and CK HIGH at T2) .................................. 79
Figure 21: Write Leveling Exit ......................................................................................................................... 80
Figure 22: CAL Timing Definition ................................................................................................................... 81
Figure 23: CAL Timing Example (Consecutive CS_n = LOW) ............................................................................ 81
Figure 24: CAL Enable Timing – tMOD_CAL ................................................................................................... 82
Figure 25: tMOD_CAL, MRS to Valid Command Timing with CAL Enabled ....................................................... 82
Figure 26: CAL Enabling MRS to Next MRS Command, tMRD_CAL .................................................................. 83
Figure 27: tMRD_CAL, Mode Register Cycle Time With CAL Enabled ............................................................... 83
Figure 28: Consecutive READ BL8, CAL3, 1tCK Preamble, Different Bank Group ............................................... 84
Figure 29: Consecutive READ BL8, CAL4, 1tCK Preamble, Different Bank Group ............................................... 84
Figure 30: Auto Self Refresh Ranges ................................................................................................................ 87
Figure 31: MPR Block Diagram ....................................................................................................................... 88
Figure 32: MPR READ Timing ........................................................................................................................ 94
Figure 33: MPR Back-to-Back READ Timing ................................................................................................... 95
Figure 34: MPR READ-to-WRITE Timing ........................................................................................................ 96
Figure 35: MPR WRITE and WRITE-to-READ Timing ...................................................................................... 97
Figure 36: MPR Back-to-Back WRITE Timing .................................................................................................. 98
Figure 37: REFRESH Timing ........................................................................................................................... 98
Figure 38: READ-to-REFRESH Timing ............................................................................................................ 99
Figure 39: WRITE-to-REFRESH Timing .......................................................................................................... 99
Figure 40: Clock Mode Change from 1/2 Rate to 1/4 Rate (Initialization) ......................................................... 102
Figure 41: Clock Mode Change After Exiting Self Refresh ................................................................................ 102
Figure 42: Comparison Between Gear-Down Disable and Gear-Down Enable ................................................. 103
Figure 43: Maximum Power-Saving Mode Entry ............................................................................................. 104
Figure 44: Maximum Power-Saving Mode Entry with PDA .............................................................................. 105
Figure 45: Maintaining Maximum Power-Saving Mode with CKE Transition ................................................... 105
Figure 46: Maximum Power-Saving Mode Exit ............................................................................................... 106
Figure 47: Command/Address Parity Operation ............................................................................................. 107
Figure 48: Command/Address Parity During Normal Operation ..................................................................... 109
Figure 49: Persistent CA Parity Error Checking Operation ............................................................................... 110
Figure 50: CA Parity Error Checking – SRE Attempt ........................................................................................ 110
Features
Figure 51: CA Parity Error Checking – SRX Attempt ........................................................................................ 111

Figure 52: CA Parity Error Checking – PDE/PDX ............................................................................................ 111
Figure 53: Parity Entry Timing Example – tMRD_PAR ..................................................................................... 112
Figure 54: Parity Entry Timing Example – tMOD_PAR ..................................................................................... 112
Figure 55: Parity Exit Timing Example – tMRD_PAR ....................................................................................... 112
Figure 56: Parity Exit Timing Example – tMOD_PAR ....................................................................................... 113
Figure 57: CA Parity Flow Diagram ................................................................................................................ 114
Figure 58: PDA Operation Enabled, BL8 ........................................................................................................ 116
Figure 59: PDA Operation Enabled, BC4 ........................................................................................................ 116
Figure 60: MRS PDA Exit ............................................................................................................................... 117
Figure 61: V REFDQ Voltage Range ................................................................................................................... 118
Figure 62: Example of V REF Set Tolerance and Step Size .................................................................................. 120
Figure 63: V REFDQ Timing Diagram for V REF,time Parameter .............................................................................. 121
Figure 64: V REFDQ Training Mode Entry and Exit Timing Diagram ................................................................... 122
Figure 65: V REF Step: Single Step Size Increment Case .................................................................................... 123
Figure 66: V REF Step: Single Step Size Decrement Case ................................................................................... 123
Figure 67: V REF Full Step: From V REF,min to V REF,maxCase .................................................................................. 124
Figure 68: V REF Full Step: From V REF,max to V REF,minCase .................................................................................. 124
Figure 69: V REFDQ Equivalent Circuit ............................................................................................................. 125
Figure 70: Connectivity Test Mode Entry ....................................................................................................... 129
Figure 71: hPPR WRA – Entry ........................................................................................................................ 133
Figure 72: hPPR WRA – Repair and Exit ......................................................................................................... 134
Figure 73: hPPR WR – Entry .......................................................................................................................... 135
Figure 74: hPPR WR – Repair and Exit ............................................................................................................ 135
Figure 75: sPPR – Entry ................................................................................................................................. 138
Figure 76: sPPR – Repair, and Exit ................................................................................................................. 138
Figure 77: tRRD Timing ................................................................................................................................ 139
Figure 78: tFAW Timing ................................................................................................................................. 140
Figure 79: REFRESH Command Timing ......................................................................................................... 141
Figure 80: Postponing REFRESH Commands (Example) ................................................................................. 142
Figure 81: Pulling In REFRESH Commands (Example) ................................................................................... 142
Figure 82: TCR Mode Example 1 ..................................................................................................................... 145
Figure 83: 4Gb with Fine Granularity Refresh Mode Example ......................................................................... 148
Figure 84: OTF REFRESH Command Timing ................................................................................................. 149
Figure 85: Self Refresh Entry/Exit Timing ...................................................................................................... 152
Figure 86: Self Refresh Entry/Exit Timing with CAL Mode ............................................................................... 153
Figure 87: Self Refresh Abort ......................................................................................................................... 154
Figure 88: Self Refresh Exit with NOP Command ............................................................................................ 155
Figure 89: Active Power-Down Entry and Exit ................................................................................................ 157
Figure 90: Power-Down Entry After Read and Read with Auto Precharge ......................................................... 158
Figure 91: Power-Down Entry After Write and Write with Auto Precharge ........................................................ 158
Figure 92: Power-Down Entry After Write ...................................................................................................... 159
Figure 93: Precharge Power-Down Entry and Exit .......................................................................................... 159
Figure 94: REFRESH Command to Power-Down Entry ................................................................................... 160
Figure 95: Active Command to Power-Down Entry ......................................................................................... 160
Figure 96: PRECHARGE/PRECHARGE ALL Command to Power-Down Entry .................................................. 161
Figure 97: MRS Command to Power-Down Entry ........................................................................................... 161
Figure 98: Power-Down Entry/Exit Clarifications – Case 1 .............................................................................. 162
Figure 99: Active Power-Down Entry and Exit Timing with CAL ...................................................................... 162
Figure 100: REFRESH Command to Power-Down Entry with CAL ................................................................... 163
Figure 101: ODT Power-Down Entry with ODT Buffer Disable Mode .............................................................. 164
Figure 102: ODT Power-Down Exit with ODT Buffer Disable Mode ................................................................. 165
Features
Figure 103: CRC Write Data Operation .......................................................................................................... 166

Figure 104: CRC Error Reporting ................................................................................................................... 175
Figure 105: CA Parity Flow Diagram .............................................................................................................. 176
Figure 106: 1tCK vs. 2tCK WRITE Preamble Mode ........................................................................................... 181
Figure 107: 1tCK vs. 2tCK WRITE Preamble Mode, tCCD = 4 ............................................................................ 182
Figure 108: 1tCK vs. 2tCK WRITE Preamble Mode, tCCD = 5 ............................................................................ 183
Figure 109: 1tCK vs. 2 tCK WRITE Preamble Mode, tCCD = 6 ........................................................................... 183
Figure 110: 1tCK vs. 2tCK READ Preamble Mode ............................................................................................ 184
Figure 111: READ Preamble Training ............................................................................................................. 185
Figure 112: WRITE Postamble ....................................................................................................................... 185
Figure 113: READ Postamble ........................................................................................................................ 186
Figure 114: Bank Group x4/x8 Block Diagram ................................................................................................ 187
Figure 115: READ Burst tCCD_S and tCCD_L Examples .................................................................................. 188
Figure 116: Write Burst tCCD_S and tCCD_L Examples ................................................................................... 188
Figure 117: tRRD Timing ............................................................................................................................... 189
Figure 118: tWTR_S Timing (WRITE-to-READ, Different Bank Group, CRC and DM Disabled) ......................... 189
Figure 119: tWTR_L Timing (WRITE-to-READ, Same Bank Group, CRC and DM Disabled) .............................. 190
Figure 120: Read Timing Definition ............................................................................................................... 192

Figure 121: Clock-to-Data Strobe Relationship .............................................................................................. 193
Figure 122: Data Strobe-to-Data Relationship ................................................................................................ 194
Figure 123: tLZ and tHZ Method for Calculating Transitions and Endpoints .................................................... 195
Figure 124: tRPRE Method for Calculating Transitions and Endpoints ............................................................. 196
Figure 125: tRPST Method for Calculating Transitions and Endpoints ............................................................. 197
Figure 126: READ Burst Operation, RL = 11 (AL = 0, CL = 11, BL8) ................................................................... 198
Figure 127: READ Burst Operation, RL = 21 (AL = 10, CL = 11, BL8) ................................................................. 199
Figure 128: Consecutive READ (BL8) with 1tCK Preamble in Different Bank Group .......................................... 200
Figure 129: Consecutive READ (BL8) with 2tCK Preamble in Different Bank Group .......................................... 200
Figure 130: Nonconsecutive READ (BL8) with 1tCK Preamble in Same or Different Bank Group ....................... 201
Figure 131: Nonconsecutive READ (BL8) with 2tCK Preamble in Same or Different Bank Group ....................... 201
Figure 132: READ (BC4) to READ (BC4) with 1tCK Preamble in Different Bank Group ...................................... 202
Figure 133: READ (BC4) to READ (BC4) with 2tCK Preamble in Different Bank Group ...................................... 202
Figure 134: READ (BL8) to READ (BC4) OTF with 1tCK Preamble in Different Bank Group ............................... 203
Figure 135: READ (BL8) to READ (BC4) OTF with 2tCK Preamble in Different Bank Group ............................... 203
Figure 136: READ (BC4) to READ (BL8) OTF with 1tCK Preamble in Different Bank Group ............................... 204
Figure 137: READ (BC4) to READ (BL8) OTF with 2tCK Preamble in Different Bank Group ............................... 204
Figure 138: READ (BL8) to WRITE (BL8) with 1 tCK Preamble in Same or Different Bank Group ........................ 205
Figure 139: READ (BL8) to WRITE (BL8) with 2 tCK Preamble in Same or Different Bank Group ........................ 205
Figure 140: READ (BC4) OTF to WRITE (BC4) OTF with 1 tCK Preamble in Same or Different Bank Group ......... 206
Figure 141: READ (BC4) OTF to WRITE (BC4) OTF with 2 tCK Preamble in Same or Different Bank Group ......... 207
Figure 142: READ (BC4) Fixed to WRITE (BC4) Fixed with 1 tCK Preamble in Same or Different Bank Group ..... 207
Figure 143: READ (BC4) Fixed to WRITE (BC4) Fixed with 2 tCK Preamble in Same or Different Bank Group ..... 208
Figure 144: READ (BC4) to WRITE (BL8) OTF with 1 tCK Preamble in Same or Different Bank Group ................ 209
Figure 145: READ (BC4) to WRITE (BL8) OTF with 2 tCK Preamble in Same or Different Bank Group ................ 209
Figure 146: READ (BL8) to WRITE (BC4) OTF with 1 tCK Preamble in Same or Different Bank Group ................ 210
Figure 147: READ (BL8) to WRITE (BC4) OTF with 2 tCK Preamble in Same or Different Bank Group ................ 210
Figure 148: READ to PRECHARGE with 1tCK Preamble .................................................................................. 211
Figure 149: READ to PRECHARGE with 2tCK Preamble .................................................................................. 212
Figure 150: READ to PRECHARGE with Additive Latency and 1tCK Preamble .................................................. 212
Figure 151: READ with Auto Precharge and 1tCK Preamble ............................................................................ 213
Figure 152: READ with Auto Precharge, Additive Latency, and 1tCK Preamble ................................................. 214
Figure 153: Consecutive READ (BL8) with 1tCK Preamble and DBI in Different Bank Group ............................ 214
Figure 154: Consecutive READ (BL8) with 1tCK Preamble and CA Parity in Different Bank Group .................... 215
Features
Figure 155: READ (BL8) to WRITE (BL8) with 1 tCK Preamble and CA Parity in Same or Different Bank Group ... 216
Figure 156: READ (BL8) to WRITE (BL8 or BC4: OTF) with 1 tCK Preamble and Write CRC in Same or Different
Bank Group ............................................................................................................................................... 217
Figure 157: READ (BC4: Fixed) to WRITE (BC4: Fixed) with 1 tCK Preamble and Write CRC in Same or Different
Bank Group ............................................................................................................................................... 218
Figure 158: Consecutive READ (BL8) with CAL (3tCK) and 1tCK Preamble in Different Bank Group .................. 218
Figure 159: Consecutive READ (BL8) with CAL (4tCK) and 1tCK Preamble in Different Bank Group .................. 219
Figure 160: Write Timing Definition .............................................................................................................. 221
Figure 161: tWPRE Method for Calculating Transitions and Endpoints ............................................................ 222
Figure 162: tWPST Method for Calculating Transitions and Endpoints ............................................................ 223
Figure 163: Rx Compliance Mask .................................................................................................................. 224
Figure 164: V CENT_DQ VREFDQ Voltage Variation .............................................................................................. 224
Figure 165: Rx Mask DQ-to-DQS Timings ...................................................................................................... 225
Figure 166: Rx Mask DQ-to-DQS DRAM-Based Timings ................................................................................. 226
Figure 167: Example of Data Input Requirements Without Training ................................................................ 227
Figure 168: WRITE Burst Operation, WL = 9 (AL = 0, CWL = 9, BL8) ................................................................. 228
Figure 169: WRITE Burst Operation, WL = 19 (AL = 10, CWL = 9, BL8) ............................................................. 229
Figure 170: Consecutive WRITE (BL8) with 1 tCK Preamble in Different Bank Group ........................................ 229
Figure 171: Consecutive WRITE (BL8) with 2 tCK Preamble in Different Bank Group ........................................ 230
Figure 172: Nonconsecutive WRITE (BL8) with 1 tCK Preamble in Same or Different Bank Group ..................... 231
Figure 173: Nonconsecutive WRITE (BL8) with 2 tCK Preamble in Same or Different Bank Group ..................... 231
Figure 174: WRITE (BC4) OTF to WRITE (BC4) OTF with 1 tCK Preamble in Different Bank Group .................... 232
Figure 175: WRITE (BC4) OTF to WRITE (BC4) OTF with 2 tCK Preamble in Different Bank Group .................... 233
Figure 176: WRITE (BC4) Fixed to WRITE (BC4) Fixed with 1 tCK Preamble in Different Bank Group ................. 233
Figure 177: WRITE (BL8) to WRITE (BC4) OTF with 1 tCK Preamble in Different Bank Group ............................ 234
Figure 178: WRITE (BC4) OTF to WRITE (BL8) with 1 tCK Preamble in Different Bank Group ............................ 235
Figure 179: WRITE (BL8) to READ (BL8) with 1 tCK Preamble in Different Bank Group ..................................... 235
Figure 180: WRITE (BL8) to READ (BL8) with 1 tCK Preamble in Same Bank Group .......................................... 236
Figure 181: WRITE (BC4) OTF to READ (BC4) OTF with 1 tCK Preamble in Different Bank Group ...................... 237
Figure 182: WRITE (BC4) OTF to READ (BC4) OTF with 1 tCK Preamble in Same Bank Group ........................... 237
Figure 183: WRITE (BC4) Fixed to READ (BC4) Fixed with 1 tCK Preamble in Different Bank Group ................. 238
Figure 184: WRITE (BC4) Fixed to READ (BC4) Fixed with 1 tCK Preamble in Same Bank Group ....................... 238
Figure 185: WRITE (BL8/BC4-OTF) to PRECHARGE with 1 tCK Preamble ........................................................ 239
Figure 186: WRITE (BC4-Fixed) to PRECHARGE with 1 tCK Preamble .............................................................. 240
Figure 187: WRITE (BL8/BC4-OTF) to Auto PRECHARGE with 1 tCK Preamble ................................................ 240
Figure 188: WRITE (BC4-Fixed) to Auto PRECHARGE with 1 tCK Preamble ...................................................... 241
Figure 189: WRITE (BL8/BC4-OTF) with 1 tCK Preamble and DBI ................................................................... 242
Figure 190: WRITE (BC4-Fixed) with 1 tCK Preamble and DBI ......................................................................... 243
Figure 191: Consecutive Write (BL8) with 1 tCK Preamble and CA Parity in Different Bank Group ..................... 244
Figure 192: Consecutive WRITE (BL8/BC4-OTF) with 1 tCK Preamble and Write CRC in Same or Different Bank
Group ....................................................................................................................................................... 245
Figure 193: Consecutive WRITE (BC4-Fixed) with 1 tCK Preamble and Write CRC in Same or Different Bank
Group ....................................................................................................................................................... 246
Figure 194: Nonconsecutive WRITE (BL8/BC4-OTF) with 1 tCK Preamble and Write CRC in Same or Different
Bank Group ............................................................................................................................................... 247
Figure 195: Nonconsecutive WRITE (BL8/BC4-OTF) with 2 tCK Preamble and Write CRC in Same or Different
Bank Group ............................................................................................................................................... 248
Figure 196: WRITE (BL8/BC4-OTF/Fixed) with 1 tCK Preamble and Write CRC in Same or Different Bank Group ... 249
Figure 197: ZQ Calibration Timing ................................................................................................................ 252
Figure 198: Functional Representation of ODT .............................................................................................. 253
Figure 199: Synchronous ODT Timing with BL8 ............................................................................................. 256
Figure 200: Synchronous ODT with BC4 ........................................................................................................ 256
Features
Figure 201: ODT During Reads ...................................................................................................................... 257

Figure 202: Dynamic ODT (1t CK Preamble; CL = 14, CWL = 11, BL = 8, AL = 0, CRC Disabled) .......................... 259
Figure 203: Dynamic ODT Overlapped with RTT(NOM) (CL = 14, CWL = 11, BL = 8, AL = 0, CRC Disabled) .......... 260
Figure 204: Asynchronous ODT Timings with DLL Off ................................................................................... 261
Figure 205: V REFDQ Voltage Range .................................................................................................................. 264
Figure 206: RESET_n Input Slew Rate Definition ............................................................................................ 267
Figure 207: Single-Ended Input Slew Rate Definition ..................................................................................... 269
Figure 208: DQ Slew Rate Definitions ............................................................................................................ 272
Figure 209: Rx Mask Relative to tDS/tDH ....................................................................................................... 274
Figure 210: Rx Mask Without Write Training .................................................................................................. 275
Figure 211: TEN Input Slew Rate Definition ................................................................................................... 276
Figure 212: CT Type-A Input Slew Rate Definition .......................................................................................... 276
Figure 213: CT Type-B Input Slew Rate Definition .......................................................................................... 277
Figure 214: CT Type-C Input Slew Rate Definition .......................................................................................... 278
Figure 215: CT Type-D Input Slew Rate Definition ......................................................................................... 278
Figure 216: Differential AC Swing and “Time Exceeding AC-Level” tDVAC ....................................................... 279
Figure 217: Single-Ended Requirements for CK .............................................................................................. 281
Figure 218: Differential Input Slew Rate Definition for CK_t, CK_c .................................................................. 282
Figure 219: V IX(CK) Definition ........................................................................................................................ 282
Figure 220: Differential Input Signal Definition for DQS_t, DQS_c .................................................................. 283
Figure 221: DQS_t, DQS_c Input Peak Voltage Calculation and Range of Exempt non-Monotonic Signaling ..... 284
Figure 222: V IXDQS Definition ........................................................................................................................ 285
Figure 223: Differential Input Slew Rate and Input Level Definition for DQS_t, DQS_c ..................................... 286
Figure 224: ADDR, CMD, CNTL Overshoot and Undershoot Definition ........................................................... 288
Figure 225: CK Overshoot and Undershoot Definition .................................................................................... 289
Figure 226: Data, Strobe, and Mask Overshoot and Undershoot Definition ..................................................... 290
Figure 227: Single-ended Output Slew Rate Definition ................................................................................... 291
Figure 228: Differential Output Slew Rate Definition ...................................................................................... 293
Figure 229: Reference Load For AC Timing and Output Slew Rate ................................................................... 294
Figure 230: Connectivity Test Mode Reference Test Load ................................................................................ 294
Figure 231: Connectivity Test Mode Output Slew Rate Definition .................................................................... 295
Figure 232: Output Driver During Connectivity Test Mode ............................................................................. 296
Figure 233: Output Driver: Definition of Voltages and Currents ...................................................................... 297
Figure 234: Alert Driver ................................................................................................................................ 301
Figure 235: ODT Definition of Voltages and Currents ..................................................................................... 302
Figure 236: ODT Timing Reference Load ....................................................................................................... 303
Figure 237: tADC Definition with Direct ODT Control .................................................................................... 305
Figure 238: tADC Definition with Dynamic ODT Control ................................................................................ 305
Figure 239: tAOFAS and tAONAS Definitions .................................................................................................. 306
Figure 240: Thermal Measurement Point ....................................................................................................... 312

Figure 241: Measurement Setup and Test Load for I DDx, IPPx, and IDDQx ........................................................... 313
Figure 242: Correlation: Simulated Channel I/O Power to Actual Channel I/O Power ....................................... 314
Features
List of Tables
Table 1: Key Timing Parameters ....................................................................................................................... 2
Table 2: Addressing ......................................................................................................................................... 2
Table 3: Ball Descriptions .............................................................................................................................. 25
Table 4: State Diagram Command Definitions ................................................................................................ 33
Table 5: Supply Power-up Slew Rate ............................................................................................................... 35
Table 6: Address Pin Mapping ........................................................................................................................ 43
Table 7: MR0 Register Definition .................................................................................................................... 43
Table 8: Burst Type and Burst Order ............................................................................................................... 45
Table 9: Address Pin Mapping ........................................................................................................................ 47
Table 10: MR1 Register Definition .................................................................................................................. 47
Table 11: Additive Latency (AL) Settings ......................................................................................................... 49
Table 12: TDQS Function Matrix .................................................................................................................... 50
Table 13: Address Pin Mapping ...................................................................................................................... 51
Table 23: Truth Table – Command .................................................................................................................. 67
Table 24: Truth Table – CKE ........................................................................................................................... 69
Table 25: MR Settings for Leveling Procedures ................................................................................................ 77
Table 26: DRAM TERMINATION Function in Leveling Mode ........................................................................... 77
Table 27: Auto Self Refresh Mode ................................................................................................................... 86
Table 28: MR3 Setting for the MPR Access Mode ............................................................................................. 88
Table 29: DRAM Address to MPR UI Translation ............................................................................................. 88
Table 30: MPR Page and MPRx Definitions ..................................................................................................... 89
Table 31: MPR Readout Serial Format ............................................................................................................. 91
Table 32: MPR Readout – Parallel Format ....................................................................................................... 92
Table 33: MPR Readout Staggered Format, x4 ................................................................................................. 93
Table 34: MPR Readout Staggered Format, x4 – Consecutive READs ................................................................ 93
Table 35: MPR Readout Staggered Format, x8 and x16 ..................................................................................... 94
Table 36: Mode Register Setting for CA Parity ................................................................................................. 109
Table 37: V REFDQ Range and Levels ................................................................................................................ 119
Table 38: V REFDQ Settings (VDDQ = 1.2V) ......................................................................................................... 125
Table 39: Connectivity Mode Pin Description and Switching Levels ................................................................ 127
Table 40: MAC Encoding of MPR Page 3 MPR3 ............................................................................................... 130
Table 41: PPR MR0 Guard Key Settings .......................................................................................................... 132
Table 42: DDR4 hPPR Timing Parameters DDR4-1600 through DDR4-3200 ..................................................... 135
Table 43: sPPR Associated Rows .................................................................................................................... 136
Table 44: PPR MR0 Guard Key Settings .......................................................................................................... 137
Table 45: DDR4 sPPR Timing Parameters DDR4-1600 through DDR4-3200 ..................................................... 138
Table 46: DDR4 Repair Mode Support Identifier ............................................................................................ 138
Table 47: Normal tREFI Refresh (TCR Enabled) .............................................................................................. 143
Table 48: MRS Definition .............................................................................................................................. 146
Table 49: REFRESH Command Truth Table .................................................................................................... 146
Table 50: tREFI and tRFC Parameters ............................................................................................................. 147
Features
Table 51: Power-Down Entry Definitions ....................................................................................................... 156

Table 52: CRC Error Detection Coverage ........................................................................................................ 167
Table 53: CRC Data Mapping for x4 Devices, BL8 ........................................................................................... 169
Table 54: CRC Data Mapping for x8 Devices, BL8 ........................................................................................... 169
Table 55: CRC Data Mapping for x16 Devices, BL8 ......................................................................................... 170
Table 56: CRC Data Mapping for x4 Devices, BC4 ........................................................................................... 170
Table 57: CRC Data Mapping for x8 Devices, BC4 ........................................................................................... 171
Table 58: CRC Data Mapping for x16 Devices, BC4 ......................................................................................... 172
Table 59: DBI vs. DM vs. TDQS Function Matrix ............................................................................................. 177
Table 60: DBI Write, DQ Frame Format (x8) ................................................................................................... 177
Table 61: DBI Write, DQ Frame Format (x16) ................................................................................................. 177
Table 62: DBI Read, DQ Frame Format (x8) .................................................................................................... 178
Table 63: DBI Read, DQ Frame Format (x16) .................................................................................................. 178
Table 64: DM vs. TDQS vs. DBI Function Matrix ............................................................................................. 179
Table 65: Data Mask, DQ Frame Format (x8) .................................................................................................. 179
Table 66: Data Mask, DQ Frame Format (x16) ................................................................................................ 179
Table 67: CWL Selection ............................................................................................................................... 182
Table 68: DDR4 Bank Group Timing Examples .............................................................................................. 187
Table 69: Read-to-Write and Write-to-Read Command Intervals .................................................................... 192
Table 70: Termination State Table ................................................................................................................. 254
Table 71: Read Termination Disable Window ................................................................................................. 254
Table 72: ODT Latency at DDR4-1600/-1866/-2133/-2400/-2666/-3200 .......................................................... 255
Table 73: Dynamic ODT Latencies and Timing (1 tCK Preamble Mode and CRC Disabled) ................................ 258
Table 74: Dynamic ODT Latencies and Timing with Preamble Mode and CRC Mode Matrix ............................ 259
Table 75: Absolute Maximum Ratings ............................................................................................................ 262
Table 76: Temperature Range ........................................................................................................................ 262
Table 77: Recommended Supply Operating Conditions .................................................................................. 263
Table 78: V DD Slew Rate ................................................................................................................................ 263
Table 79: Leakages ....................................................................................................................................... 264
Table 80: V REFDQ Specification ...................................................................................................................... 265
Table 81: V REFDQ Range and Levels ................................................................................................................ 266
Table 82: RESET_n Input Levels (CMOS) ....................................................................................................... 267
Table 83: Command and Address Input Levels: DDR4-1600 Through DDR4-2400 ........................................... 267
Table 84: Command and Address Input Levels: DDR4-2666 ............................................................................ 268
Table 85: Command and Address Input Levels: DDR4-2933 and DDR4-3200 ................................................... 268
Table 86: Single-Ended Input Slew Rates ....................................................................................................... 269
Table 87: Command and Address Setup and Hold Values Referenced – AC/DC-Based ..................................... 270
Table 88: Derating Values for tIS/tIH – AC100DC75-Based .............................................................................. 270
Table 89: Derating Values for tIS/tIH – AC90/DC65-Based .............................................................................. 271
Table 90: DQ Input Receiver Specifications .................................................................................................... 272
Table 91: Rx Mask and tDS/tDH without Write Training .................................................................................. 275
Table 92: TEN Input Levels (CMOS) .............................................................................................................. 275
Table 93: CT Type-A Input Levels .................................................................................................................. 276
Table 94: CT Type-B Input Levels .................................................................................................................. 277
Table 95: CT Type-C Input Levels (CMOS) ..................................................................................................... 277
Table 96: CT Type-D Input Levels .................................................................................................................. 278
Table 97: Differential Input Swing Requirements for CK_t, CK_c ..................................................................... 279
Table 98: Minimum Time AC Time tDVAC for CK ........................................................................................... 280
Table 99: Single-Ended Requirements for CK ................................................................................................. 281
Table 100: CK Differential Input Slew Rate Definition ..................................................................................... 281
Table 101: Cross Point Voltage For CK Differential Input Signals at DDR4-1600 through DDR4-2400 ................ 283
Table 102: Cross Point Voltage For CK Differential Input Signals at DDR4-2666 through DDR4-3200 ................ 283
Features
Table 103: DDR4-1600 through DDR4-2400 Differential Input Swing Requirements for DQS_t, DQS_c ............. 283
Table 104: DDR4-2633 through DDR4-3200 Differential Input Swing Requirements for DQS_t, DQS_c ............. 284
Table 105: Cross Point Voltage For Differential Input Signals DQS ................................................................... 285
Table 106: DQS Differential Input Slew Rate Definition .................................................................................. 286
Table 107: DDR4-1600 through DDR4-2400 Differential Input Slew Rate and Input Levels for DQS_t, DQS_c ... 286
Table 108: DDR4-2666 through DDR4-3200 Differential Input Slew Rate and Input Levels for DQS_t, DQS_c ... 287
Table 109: ADDR, CMD, CNTL Overshoot and Undershoot/Specifications ...................................................... 288
Table 110: CK Overshoot and Undershoot/ Specifications .............................................................................. 289
Table 111: Data, Strobe, and Mask Overshoot and Undershoot/ Specifications ................................................ 290
Table 112: Single-Ended Output Levels ......................................................................................................... 290
Table 113: Single-Ended Output Slew Rate Definition .................................................................................... 291
Table 114: Single-Ended Output Slew Rate .................................................................................................... 292
Table 115: Differential Output Levels ............................................................................................................. 292
Table 116: Differential Output Slew Rate Definition ....................................................................................... 292
Table 117: Differential Output Slew Rate ....................................................................................................... 293
Table 118: Connectivity Test Mode Output Levels .......................................................................................... 294
Table 119: Connectivity Test Mode Output Slew Rate ..................................................................................... 295
Table 120: Output Driver Electrical Characteristics During Connectivity Test Mode ......................................... 297
Table 121: Strong Mode (34˖) Output Driver Electrical Characteristics ........................................................... 298
Table 122: Weak Mode (48˖) Output Driver Electrical Characteristics ............................................................. 299
Table 123: Output Driver Sensitivity Definitions ............................................................................................ 300
Table 124: Output Driver Voltage and Temperature Sensitivity ....................................................................... 300
Table 125: Alert Driver Voltage ...................................................................................................................... 301
Table 126: ODT DC Characteristics ............................................................................................................... 302
Table 127: ODT Sensitivity Definitions .......................................................................................................... 303
Table 128: ODT Voltage and Temperature Sensitivity ..................................................................................... 303
Table 129: ODT Timing Definitions ............................................................................................................... 304
Table 130: Reference Settings for ODT Timing Measurements ........................................................................ 304
Table 131: DRAM Package Electrical Specifications for x4 and x8 Devices ....................................................... 307
Table 132: DRAM Package Electrical Specifications for x16 Devices ................................................................ 308
Table 133: Pad Input/Output Capacitance ..................................................................................................... 310
Table 134: Thermal Characteristics ............................................................................................................... 311
Table 135: Basic IDD, IPP, and IDDQ Measurement Conditions .......................................................................... 314
Table 136: IDD0 and IPP0 Measurement-Loop Pattern1 .................................................................................... 318
Table 137: IDD1 Measurement – Loop Pattern1 ............................................................................................... 319
Table 138: IDD2N, IDD3N, and IPP3P Measurement – Loop Pattern1 .................................................................... 320
Table 139: IDD2NT Measurement – Loop Pattern1 ............................................................................................ 321
Table 140: IDD4R Measurement – Loop Pattern1 .............................................................................................. 322
Table 141: IDD4W Measurement – Loop Pattern1 ............................................................................................. 323
Table 142: IDD4Wc Measurement – Loop Pattern1 ............................................................................................ 324
Table 143: IDD5R Measurement – Loop Pattern1 .............................................................................................. 325
Table 144: IDD7 Measurement – Loop Pattern1 ............................................................................................... 326
Table 145: Timings used for I DD, IPP, and IDDQ Measurement – Loop Patterns .................................................. 327
Table 146: IDD, IPP, and IDDQ Current Limits – Rev. B (0°C ื T C ื 95°C) ............................................................. 328
Table 147: IDD, IPP, and IDDQ Current Limits – Rev. B (0°C ื T C ื 105°C) ........................................................... 329
Table 148: IDD, IPP, and IDDQ Current Limits – Rev. B (0°C ื T C ื 125°C) ........................................................... 331
Table 149: IDD, IPP, and IDDQ Current Limits – Rev. F(0°C ื T C ื 95°C) .............................................................. 332
Table 150: IDD, IPP, and IDDQ Current Limits – Rev. F(0°C ื T C ื 105°C) ............................................................ 334
Table 151: IDD, IPP, and IDDQ Current Limits – Rev. F(0°C ื T C ื 125°C) ............................................................ 336
Table 152: Backward Compatibility ............................................................................................................... 339
Table 153: DDR4-1600 Speed Bins and Operating Conditions ......................................................................... 341
Features
Table 160: Refresh Parameters by Device Density ........................................................................................... 357
Table 161: Electrical Characteristics and AC Timing Parameters ..................................................................... 358
Table 162: Electrical Characteristics and AC Timing Parameters ..................................................................... 370
Table 163: Options – Speed Based ................................................................................................................. 383
Table 164: Options – Width Based ................................................................................................................. 384
Important Notes and Warnings
Important Notes and Warnings

Micron Technology, Inc. ("Micron") reserves the right to make changes to information published in this document,
including without limitation specifications and product descriptions. This document supersedes and replaces all
information supplied prior to the publication hereof. You may not rely on any information set forth in this docu-
ment if you obtain the product described herein from any unauthorized distributor or other source not authorized
by Micron.
Automotive Applications. Products are not designed or intended for use in automotive applications unless specifi-
cally designated by Micron as automotive-grade by their respective data sheets. Distributor and customer/distrib-
utor shall assume the sole risk and liability for and shall indemnify and hold Micron harmless against all claims,
costs, damages, and expenses and reasonable attorneys' fees arising out of, directly or indirectly, any claim of
product liability, personal injury, death, or property damage resulting directly or indirectly from any use of non-
automotive-grade products in automotive applications. Customer/distributor shall ensure that the terms and con-
ditions of sale between customer/distributor and any customer of distributor/customer (1) state that Micron
products are not designed or intended for use in automotive applications unless specifically designated by Micron
as automotive-grade by their respective data sheets and (2) require such customer of distributor/customer to in-
demnify and hold Micron harmless against all claims, costs, damages, and expenses and reasonable attorneys'
fees arising out of, directly or indirectly, any claim of product liability, personal injury, death, or property damage
resulting from any use of non-automotive-grade products in automotive applications.
Critical Applications. Products are not authorized for use in applications in which failure of the Micron compo-
nent could result, directly or indirectly in death, personal injury, or severe property or environmental damage
("Critical Applications"). Customer must protect against death, personal injury, and severe property and environ-
mental damage by incorporating safety design measures into customer's applications to ensure that failure of the
Micron component will not result in such harms. Should customer or distributor purchase, use, or sell any Micron
component for any critical application, customer and distributor shall indemnify and hold harmless Micron and
its subsidiaries, subcontractors, and affiliates and the directors, officers, and employees of each against all claims,
costs, damages, and expenses and reasonable attorneys' fees arising out of, directly or indirectly, any claim of
product liability, personal injury, or death arising in any way out of such critical application, whether or not Mi-
cron or its subsidiaries, subcontractors, or affiliates were negligent in the design, manufacture, or warning of the
Micron product.
Customer Responsibility. Customers are responsible for the design, manufacture, and operation of their systems,
applications, and products using Micron products. ALL SEMICONDUCTOR PRODUCTS HAVE INHERENT FAIL-
URE RATES AND LIMITED USEFUL LIVES. IT IS THE CUSTOMER'S SOLE RESPONSIBILITY TO DETERMINE
WHETHER THE MICRON PRODUCT IS SUITABLE AND FIT FOR THE CUSTOMER'S SYSTEM, APPLICATION, OR
PRODUCT. Customers must ensure that adequate design, manufacturing, and operating safeguards are included
in customer's applications and products to eliminate the risk that personal injury, death, or severe property or en-
vironmental damages will result from failure of any semiconductor component.
Limited Warranty. In no event shall Micron be liable for any indirect, incidental, punitive, special or consequential
damages (including without limitation lost profits, lost savings, business interruption, costs related to the removal
or replacement of any products or rework charges) whether or not such damages are based on tort, warranty,
breach of contract or other legal theory, unless explicitly stated in a written agreement executed by Micron's duly
authorized representative.
General Notes and Description

Description
The DDR4 SDRAM is a high-speed dynamic random-access memory internally config-
ured as an eight-bank DRAM for the x16 configuration and as a 16-bank DRAM for the
x8 configurations. The DDR4 SDRAM uses an 8n-prefetch architecture to achieve high-

speed operation. The 8n-prefetch architecture is combined with an interface designed
to transfer two data words per clock cycle at the I/O pins.
A single READ or WRITE operation for the DDR4 SDRAM consists of a single 8n-bit
wide, four-clock data transfer at the internal DRAM core and two corresponding n-bit
wide, one-half-clock-cycle data transfers at the I/O pins.
Industrial Temperature
An industrial temperature (IT) device option requires that the case temperature not ex-
ceed below –40°C or above 95°C. JEDEC specifications require the refresh rate to double
when T C exceeds 85°C; this also requires use of the high-temperature self refresh option.
Additionally, ODT resistance and the input/output impedance must be derated when
operating temperature < 0°C.
Automotive Temperature
The automotive temperature (AT) device option requires that the case temperature not
exceed below –40°C or above 105°C. The specifications require the refresh rate to 2X
when T C exceeds 85°C; 4X when T C exceeds 95°C. Additionally, ODT resistance and the
input/output impedance must be derated when operating temperature < 0°C.
Ultra-high Temperature
The ultra-high temperature (UT) device option requires that the case temperature not
exceed below –40°C or above 125°C. The specifications require the refresh rate to 2X
when T C exceeds 85°C; 4X when T C exceeds 95°C, 8X when T C exceeds 105°C. Addition-
ally, ODT resistance and the input/output impedance must be derated when operating
temperature < 0°C.
General Notes
• The functionality and the timing specifications discussed in this data sheet are for the
DLL enable mode of operation (normal operation), unless specifically stated other-
wise.
• Throughout the data sheet, the various figures and text refer to DQs as "DQ." The DQ
term is to be interpreted as any and all DQ collectively, unless specifically stated oth-
erwise.
• The terms "_t" and "_c" are used to represent the true and complement of a differen-
tial signal pair. These terms replace the previously used notation of "#" and/or over-
bar characters. For example, differential data strobe pair DQS, DQS# is now referred
to as DQS_t, DQS_c.
• The term "_n" is used to represent a signal that is active LOW and replaces the previ-
ously used "#" and/or overbar characters. For example: CS# is now referred to as
CS_n.
• The terms "DQS" and "CK" found throughout the data sheet are to be interpreted as
DQS_t and DQS_c, and CK_t and CK_c respectively, unless specifically stated other-
wise.
• Complete functionality may be described throughout the entire document; any page
or diagram may have been simplified to convey a topic and may not be inclusive of all
requirements.
• Any specific requirement takes precedence over a general statement.

• Any functionality not specifically stated here within is considered undefined, illegal,
and not supported, and can result in unknown operation.
• Addressing is denoted as BG[n] for bank group, BA[n] for bank address, and A[n] for
row/col address.
• The NOP command is not allowed, except when exiting maximum power savings
mode or when entering gear-down mode, and only a DES command should be used.
• Not all features described within this document may be available on the rev. A (first)
version.
• Not all specifications listed are finalized industry standards; best conservative esti-
mates have been provided when an industry standard has not been finalized.
• Although it is implied throughout the specification, the DRAM must be used after V DD
has reached the stable power-on level, which is achieved by toggling CKE at least once
every 8192 × tREFI. However, in the event CKE is fixed HIGH, toggling CS_n at least
once every 8192 × tREFI is an acceptable alternative. Placing the DRAM into self re-
fresh mode also alleviates the need to toggle CKE.
• Not all features designated in the data sheet may be supported by earlier die revisions
due to late definition by JEDEC.
• A x16 device's DQ bus is comprised of two bytes. If only one of the bytes needs to be
used, use the lower byte for data transfers and terminate the upper byte as noted:
– Connect UDQS_t to V DDQ or V SS/VSSQ via a resistor in the 200വ range.
– Connect UDQS_c to the opposite rail via a resistor in the same 200വ range.
– Connect UDM to V DDQ via a large (10,000വ) pull-up resistor.
– Connect UDBI to V DDQ via a large (10,000വ) pull-up resistor.
– Connect DQ [15:8] individually to V DDQ via a large (10,000വ) resistors, or float DQ
[15:8].
Definitions of the Device-Pin Signal Level

• HIGH: A device pin is driving the logic 1 state.
• LOW: A device pin is driving the logic 0 state.
• High-Z: A device pin is tri-state.
• ODT: A device pin terminates with the ODT setting, which could be terminating or tri-
state depending on the mode register setting.
Definitions of the Bus Signal Level

• HIGH: One device on the bus is HIGH, and all other devices on the bus are either ODT
or High-Z. The voltage level on the bus is nominally V DDQ.
• LOW: One device on the bus is LOW, and all other devices on the bus are either ODT
or High-Z. The voltage level on the bus is nominally V OL(DC) if ODT was enabled, or
VSSQ if High-Z.
• High-Z: All devices on the bus are High-Z. The voltage level on the bus is undefined as
the bus is floating.
• ODT: At least one device on the bus is ODT, and all others are High-Z. The voltage lev-
el on the bus is nominally V DDQ.
Functional Block Diagrams
Functional Block Diagrams

DDR4 SDRAM is a high-speed, CMOS dynamic random access memory. It is internally
configured as an 16-bank (4-banks per Bank Group) DRAM.
Figure 2: 512 Meg x 8 Functional Block Diagram
%DQN %DQN
%DQN &5&DQG $/(57
2'7 7R2'7RXWSXWGULYHUV %DQN
%DQN %DQN SDULW\FRQWURO 9''4
%DQN %DQN
=4&$/ %* %DQN*URXS
7R=4&RQWURO
%DQN %DQN
5(6(7BQ 9UHI'4 %DQN
%DQN =4
&RQWURO %& %DQN %DQN FRQWURO
&.( ORJLF 27) %DQN %DQN
%* %DQN*URXS
&.BW&.BF &5&
2'7
3DULW\ %DQN FRQWURO =4
3$5 %DQN
%DQN %DQN HUV
6HQVHDPSOLILHUV
%DQN %DQN
7(1 $$
%DQN %DQN 9''4

%* %DQN*URXS &.BW&.BF
&6BQ &RPPDQGGHFRGH
$$$ %DQN %DQN 577S 577Q 577Z
$&7BQ 5$6BQ&$6BQ:(BQ %DQN %DQN 6HQVHDPSOLILHUV
%DQN %DQN
%DQN '//
0RGHUHJLVWHUV %DQN
%* %DQN*URXS

5RZ
DGGUHVV 0HPRU\
6HQVHDPSOLILHUV &ROXPQV '4>@
5RZ ODWFK DUUD\ [
DGGUHVV [[ DQG 5HDG '4>@
DQG '46BW'46BF
08; GHFRGHU GULYHUV

5HIUHVK
FRXQWHU
6HQVHDPSOLILHUV 5($' 9''4
),)2
DQG 577S 577Q 577Z
%& '%,
GDWD
%* 08;
%&
DQG 27) &5&
%$
,2JDWLQJ *OREDO
FRQWURO [ :ULWH
'0PDVNORJLF ,2JDWLQJ
$>@ ORJLF GULYHUV
%$>@ $GGUHVV &.BW&.BF '46BW
UHJLVWHU DQG
%*>@ LQSXW '46BF
ORJLF 9''4

[
577S 577Q 577Z

'DWD
&ROXPQ &ROXPQ LQWHUIDFH &ROXPQ
DGGUHVV GHFRGHU %&QLEEOH
FRXQWHU
ODWFK
9UHI'4
7'46BF
&ROXPQVDQG
'%,BQ
'0BQ
7'46BW
Figure 3: 256 Meg x 16 Functional Block Diagram
&5&DQG $/(57
2'7 7R2'7RXWSXWGULYHUV SDULW\FRQWURO 9''4
=4&$/ 7R=4&RQWURO
5(6(7BQ 9UHI'4 =4
&RQWURO %& FRQWURO
&.( ORJLF 27)
&.BW&.BF &5&
2'7
3DULW\ %DQN FRQWURO =4
3$5 %DQN
%DQN %DQN
%DQN %DQN
7(1 $$
%DQN %DQN 9''4
%* %DQN*URXS &.BW&.BF
&6BQ &RPPDQGGHFRGH
$$$ %DQN %DQN 577S 577Q 577Z
$&7BQ 5$6BQ&$6BQ:(BQ %DQN %DQN
%DQN %DQN
%DQN '//
0RGHUHJLVWHUV %DQN
%* %DQN*URXS

5RZ
DGGUHVV 0HPRU\ '4>@
6HQVHDPSOLILHUV &ROXPQV '4>@
5RZ ODWFK DUUD\
DGGUHVV [[ DQG 5HDG
DQG /'46BW/'46BF8'46BW8'46BF '4>@
08; GHFRGHU GULYHUV

5HIUHVK
FRXQWHU
6HQVHDPSOLILHUV 9''4
5($'
),)2
DQG %& '%, 577S 577Q 577Z

%* GDWD
%&
DQG 08;
27) &5&
%$ /'46BW
,2JDWLQJ *OREDO
FRQWURO [ :ULWH /'46BF
'0PDVNORJLF ,2JDWLQJ
$>@ ORJLF GULYHUV
%$>@ $GGUHVV &.BW&.BF
DQG 8'46BW
%*>@ UHJLVWHU
LQSXW 8'46BF
ORJLF 9''4

[
577S 577Q 577Z

'DWD
&ROXPQ &ROXPQ LQWHUIDFH &ROXPQ
DGGUHVV GHFRGHU %&QLEEOH
FRXQWHU
ODWFK /'%,BQ
9UHI'4
/'0BQ
&ROXPQVDQG
8'%,BQ
8'0BQ
Ball Assignments
Ball Assignments
Figure 4: 78-Ball x4, x8 Ball Assignments

$ $
9'' 9664 1)1) 1)1)'0BQ 9664 966
7'46BF '%,BQ7'46BW
% %
933 9''4 '46BF '4 9''4 =4
& &
9''4 '4 '46BW 9'' 966 9''4
' '
9664 1)'4 '4 '4 1)'4 9664
( (
966 9''4 1)'4 1)'4 9''4 966
) )
9'' &2'7 2'7 &.BW &.BF 9''
* *
966 &&.( &.( &6BQ &&6BQ 7(11)
+ +
9'' :(BQ &$6BQ 5$6BQ
$&7BQ $ $
966
$
- -
95()&$ %* $$3 $%&BQ %* 9''
. .
966 %$ $ $ %$ 966
/ /
5(6(7BQ $ $ $ $ $/(57BQ
0 0
9'' $ $ $ $ 933
1 1
966 $ 3$5 $1)1& $ 9''
1)1&
Notes: 1. See Ball Descriptions.

2. A comma “,” separates the configuration; a slash “/” defines a mode register selectable
function, command/address function, density, or package dependence.
3. Address bits (including bank groups) are density- and configuration-dependent (see Ad-
dressing).
Ball Assignments
Figure 5: 96-Ball x16 Ball Assignments

$ $
9''4 9664 '4 8'46BF 9664 9''4
% %
933 966 9'' 8'46BW '4 9''
& &
9''4 '4 '4 '4 '4 9664
' '
9'' 9664 '4 '4 9664 9''4
( (
1)8'0BQ 1)/'0BQ
966 8'%,BQ
9664 /'%,BQ 9664 966
) )
9664 9''4 /'46BF '4 9''4 =4
* *
9''4 '4 /'46BW 9'' 966 9''4
+ +
9664 '4 '4 '4 '4 9664
- -
9'' 9''4 '4 '4 9''4 9''
. .
966 &.( 2'7 &.BW &.BF 966
/ /
:(BQ 5$6BQ
9'' $
$&7BQ &6BQ $ 9''
0 0
95()&$ %* $$3 $%&BQ &$6BQ
$
966
1 1
966 %$ $ $ %$ 7(1
3 3
5(6(7BQ $ $ $ $ $/(57BQ
5 5
9'' $ $ $ $ 933
7 7
966 $ 3$5 1)1& $ 9''
Notes: 1. See Ball Descriptions.

2. A slash “/” defines a mode register selectable function, command/address function, den-
sity, or package dependence.
3. Address bits (including bank groups) are density- and configuration-dependent (see Ad-
dressing).
Ball Descriptions
Ball Descriptions
The pin description table below is a comprehensive list of all possible pins for DDR4 de-
vices. All pins listed may not be supported on the device defined in this data sheet. See
the Ball Assignments section to review all pins used on this device.
Table 3: Ball Descriptions
Symbol Type Description

A[17:0] Input Address inputs: Provide the row address for ACTIVATE commands and the column
address for READ/WRITE commands to select one location out of the memory array in
the respective bank. (A10/AP, A12/BC_n, WE_n/A14, CAS_n/A15, RAS_n/A16 have addi-
tional functions, see individual entries in this table.) The address inputs also provide
the op-code during the MODE REGISTER SET command. A16 is used on some 8Gb and
16Gb parts. A17 connection is part-number specific; Contact vendor for more infor-
mation.
A10/AP Input Auto precharge: A10 is sampled during READ and WRITE commands to determine
whether auto precharge should be performed to the accessed bank after a READ or
WRITE operation. (HIGH = auto precharge; LOW = no auto precharge.) A10 is sam-
pled during a PRECHARGE command to determine whether the PRECHARGE applies
to one bank (A10 LOW) or all banks (A10 HIGH). If only one bank is to be precharged,
the bank is selected by the bank group and bank addresses.
A12/BC_n Input Burst chop: A12/BC_n is sampled during READ and WRITE commands to determine if
burst chop (on-the-fly) will be performed. (HIGH = no burst chop; LOW = burst chop-
ped). See the Command Truth Table.
ACT_n Input Command input: ACT_n indicates an ACTIVATE command. When ACT_n (along with
CS_n) is LOW, the input pins RAS_n/A16, CAS_n/A15, and WE_n/A14 are treated as
row address inputs for the ACTIVATE command. When ACT_n is HIGH (along with
CS_n LOW), the input pins RAS_n/ A16, CAS_n/A15, and WE_n/A14 are treated as nor-
mal commands that use the RAS_n, CAS_n, and WE_n signals. See the Command
Truth Table.
BA[1:0] Input Bank address inputs: Define the bank (within a bank group) to which an ACTIVATE,
READ, WRITE, or PRECHARGE command is being applied. Also determines which
mode register is to be accessed during a MODE REGISTER SET command.
BG[1:0] Input Bank group address inputs: Define the bank group to which an ACTIVATE, READ,
WRITE, or PRECHARGE command is being applied. Also determines which mode regis-
ter is to be accessed during a MODE REGISTER SET command. BG[1:0] are used in the
x4 and x8 configurations. BG1 is not used in the x16 configuration.
C0/CKE1, Input Stack address inputs: These inputs are used only when devices are stacked; that is,
C1/CS1_n, they are used in 2H, 4H, and 8H stacks for x4 and x8 configurations (these pins are
C2/ODT1 not used in the x16 configuration). DDR4 will support a traditional DDP package,
which uses these three signals for control of the second die (CS1_n, CKE1, ODT1).
DDR4 is not expected to support a traditional QDP package. For all other stack con-
figurations, such as a 4H or 8H, it is assumed to be a single-load (master/slave) type of
configuration where C0, C1, and C2 are used as chip ID selects in conjunction with a
single CS_n, CKE, and ODT signal.
CK_t, Input Clock: Differential clock inputs. All address, command, and control input signals are
CK_c sampled on the crossing of the positive edge of CK_t and the negative edge of CK_c.
Ball Descriptions
Table 3: Ball Descriptions (Continued)

CKE Input Clock enable: CKE HIGH activates and CKE LOW deactivates the internal clock sig-
nals, device input buffers, and output drivers. Taking CKE LOW provides PRECHARGE
POWER-DOWN and SELF REFRESH operations (all banks idle), or active power-down
(row active in any bank). CKE is asynchronous for self refresh exit, however, timing
parameters such as tXS are still calculated from the first rising clock edge where CKE
HIGH satisfies tIS. After VREFCA has become stable during the power-on and initializa-
tion sequence, it must be maintained during all operations (including SELF REFRESH).
CKE must be maintained HIGH throughout read and write accesses. Input buffers (ex-
cluding CK_t, CK_c, ODT, RESET_n, and CKE) are disabled during power-down. Input
buffers (excluding CKE and RESET_n) are disabled during self refresh.
CS_n Input Chip select: All commands are masked when CS_n is registered HIGH. CS_n provides
for external rank selection on systems with multiple ranks. CS_n is considered part of
the command code.
DM_n, Input Input data mask: DM_n is an input mask signal for write data. Input data is masked
UDM_n when DM is sampled LOW coincident with that input data during a write access. DM
LDM_n is sampled on both edges of DQS. DM is not supported on x4 configurations. The
UDM_n and LDM_n pins are used in the x16 configuration: UDM_n is associated with
DQ[15:8]; LDM_n is associated with DQ[7:0]. The DM, DBI, and TDQS functions are en-
abled by mode register settings. See the Data Mask section.
ODT Input On-die termination: ODT (registered HIGH) enables termination resistance internal
to the DDR4 SDRAM. When enabled, ODT (RTT) is applied only to each DQ, DQS_t,
DQS_c, DM_n/DBI_n/TDQS_t, and TDQS_c signal for the x4 and x8 configurations
(when the TDQS function is enabled via mode register). For the x16 configuration, RTT
is applied to each DQ, UDQS_t, UDQS_c, LDQS_t, LDQS_c, UDM_n, and LDM_n signal.
The ODT pin will be ignored if the mode registers are programmed to disable RTT.
PAR Input Parity for command and address: This function can be enabled or disabled via the
mode register. When enabled, the parity signal covers all command and address in-
puts, including ACT_n, RAS_n/A16, CAS_n/A15, WE_n/A14, A[17:0], A10/AP, A12/BC_n,
BA[1:0], and BG[1:0] with C0, C1, and C2 on 3DS only devices. Control pins NOT cov-
ered by the parity signal are CS_n, CKE, and ODT. Unused address pins that are densi-
ty- and configuration-specific should be treated internally as 0s by the DRAM parity
logic. Command and address inputs will have parity check performed when com-
mands are latched via the rising edge of CK_t and when CS_n is LOW.
RAS_n/A16, Input Command inputs: RAS_n/A16, CAS_n/A15, and WE_n/A14 (along with CS_n and
CAS_n/A15, ACT_n) define the command and/or address being entered. See the ACT_n descrip-
WE_n/A14 tion in this table.
RESET_n Input Active LOW asynchronous reset: Reset is active when RESET_n is LOW, and inac-
tive when RESET_n is HIGH. RESET_n must be HIGH during normal operation. RESET_n
is a CMOS rail-to-rail signal with DC HIGH and LOW at 80% and 20% of VDD (960 mV
for DC HIGH and 240 mV for DC LOW).
TEN Input Connectivity test mode: TEN is active when HIGH and inactive when LOW. TEN
must be LOW during normal operation. TEN is a CMOS rail-to-rail signal with DC
HIGH and LOW at 80% and 20% of VDD (960mV for DC HIGH and 240mV for DC
LOW). On Micron 3DS devices, connectivity test mode is not supported and the TEN
pin should be considered NF maintained LOW at all times.
Ball Descriptions
Table 3: Ball Descriptions (Continued)

DQ I/O Data input/output: Bidirectional data bus. DQ represents DQ[3:0], DQ[7:0], and
DQ[15:0] for the x4, x8, and x16 configurations, respectively. If write CRC is enabled
via mode register, the write CRC code is added at the end of data burst. Any one or
all of DQ0, DQ1, DQ2, and DQ3 may be used to monitor the internal VREF level during
test via mode register setting MR[4] A[4] = HIGH, training times change when ena-
bled. During this mode, the RTT value should be set to High-Z. This measurement is
for verification purposes and is NOT an external voltage supply pin.
DBI_n, I/O DBI input/output: Data bus inversion. DBI_n is an input/output signal used for data
UDBI_n, bus inversion in the x8 configuration. UDBI_n and LDBI_n are used in the x16 configu-
LDBI_n ration; UDBI_n is associated with DQ[15:8], and LDBI_n is associated with DQ[7:0]. The
DBI feature is not supported on the x4 configuration. DBI is not supported for 3DS
devices and should be disabled in MR5. DBI can be configured for both READ (out-
put) and WRITE (input) operations depending on the mode register settings. The DM,
DBI, and TDQS functions are enabled by mode register settings. See the Data Bus In-
version section.
DQS_t, I/O Data strobe: Output with READ data, input with WRITE data. Edge-aligned with
DQS_c, READ data, centered-aligned with WRITE data. For the x16, LDQS corresponds to the
UDQS_t, data on DQ[7:0]; UDQS corresponds to the data on DQ[15:8]. For the x4 and x8 con-
UDQS_c, figurations, DQS corresponds to the data on DQ[3:0] and DQ[7:0], respectively. DDR4
LDQS_t, SDRAM supports a differential data strobe only and does not support a single-ended
LDQS_c data strobe.
ALERT_n Output Alert output: This signal allows the DRAM to indicate to the system's memory con-
troller that a specific alert or event has occurred. Alerts will include the command/
address parity error and the CRC data error when either of these functions is enabled
in the mode register.
TDQS_t, Output Termination data strobe: TDQS_t and TDQS_c are used by x8 DRAMs only. When
TDQS_c enabled via the mode register, the DRAM will enable the same RTT termination resist-
ance on TDQS_t and TDQS_c that is applied to DQS_t and DQS_c. When the TDQS
function is disabled via the mode register, the DM/TDQS_t pin will provide the DATA
MASK (DM) function, and the TDQS_c pin is not used. The TDQS function must be dis-
abled in the mode register for both the x4 and x16 configurations. The DM function
is supported only in x8 and x16 configurations.
VDD Supply Power supply: 1.2V ±0.060V.
VDDQ Supply DQ power supply: 1.2V ±0.060V.
VPP Supply DRAM activating power supply: 2.5V –0.125V/+0.250V.
VREFCA Supply Reference voltage for control, command, and address pins.
VSS Supply Ground.
VSSQ Supply DQ ground.
ZQ Reference Reference ball for ZQ calibration: This ball is tied to an external 240˖ resistor
(RZQ), which is tied to VSSQ.
RFU – Reserved for future use.
NC – No connect: No internal electrical connection is present.
NF – No function: Internal connection is present but has no function.
Package Dimensions
Package Dimensions
Figure 6: 78-Ball FBGA – ×4, ×8, "RH"
0.155
Seating plane
A 0.12 A
1.8 CTR
nonconductive
overmold
78X Ø0.47
Dimensions apply
to solder balls post- Ball A1 ID Ball A1 ID
reflow on Ø0.42 SMD (covered by SR)
ball pads. 9 8 7 3 2 1
A
B
C
D
E
10.5 ±0.1 F
9.6 CTR G
H
J
K
L
M
0.8 TYP N
0.8 TYP 1.1 ±0.1
6.4 CTR 0.34 ±0.05
9 ±0.1
Notes: 1. All dimensions are in millimeters.

2. Solder ball material: SAC302 (Pb-free 96.8% Sn, 3% Ag, 0.2% Cu).
3. Reference CSN33 for recommended PCB pad dimension for this package.
Package Dimensions
Figure 7: 78-Ball FBGA – ×4, ×8, "SA"
0.155
Seating plane
A 0.12 A
1.8 CTR
nonconductive
overmold
78X Ø0.47
Dimensions apply
ball pads.
9 8 7 3 2 1
A
B
C
D
E
11 ±0.1 F
9.6 CTR G
H
J
K
L
M
0.8 TYP N
0.8 TYP 1.1 ±0.1
6.4 CTR 0.34 ±0.05
7.5 ±0.1

Package Dimensions
Figure 8: 96-Ball FBGA – ×16, "GE"
0.155
Seating plane
A 0.12 A
1.8 CTR
Nonconductive
overmold
96X Ø0.47
Dimensions apply
ball pads. 9 8 7 3 2 1
A
B
C
D
E
F
14 ±0.1 G
H
12 CTR J
K
L
M
N
P
R
0.8 TYP T
0.8 TYP 1.1 ±0.1
6.4 CTR 0.29 MIN
9 ±0.1

Package Dimensions
Figure 9: 96-Ball FBGA – ×16, "LY"
0.155
Seating plane
A 0.12 A
1.8 CTR
Nonconductive
overmold
96X Ø0.47
Dimensions apply
reflow on Ø0.42 (covered by SR)
SMD ball pads.
9 8 7 3 2 1
A
B
C
D
E
F
13.5 ±0.1 G
H
12 CTR J
K
L
M
N
P
R
0.8 TYP T
0.8 TYP 1.1 ±0.1
6.4 CTR 0.34 ±0.05
7.5 ±0.1

State Diagram
State Diagram
This simplified state diagram provides an overview of the possible state transitions and
the commands to control them. Situations involving more than one bank, the enabling
or disabling of on-die termination, and some other events are not captured in full de-
tail.
Figure 10: Simplified State Diagram
IVREFDQ,
RTT, and
MPSM
so on
RESET SRX* = SRX with NOP
From any state
SRX*
CKE_L
MRS
Power MRS
SRX* MRS, MPR,
applied Reset write leveling, Self
Power-On RESET Initialization PDA VREFDQ training
procedure refresh
mode
TEN = 1
TEN = 1 MRS SRX
ZQCL
MRS MRS
SRE
MRS
Connectivity ZQ
test calibration
ZQCL,ZQCS
Idle
REF
Refreshing
TEN = 0
RESET
ACT PDE
CKE_L CKE_L
PDX
Active
Precharge
power- Activating power-
down
down
PDX
PDE
Bank
active
WRITE WRITE READ READ
WRITE A READ A
READ
Writing Reading
WRITE
WRITE A READ A
WRITE A READ A
PRE, PREA
Writing PRE, PREA PRE, PREA
Reading
Precharging
Automatic
sequence
Command
sequence
State Diagram
Table 4: State Diagram Command Definitions
Command Description
ACT Active
MPR Multipurpose register
MRS Mode register set
PDE Enter power-down
PDX Exit power-down
PRE Precharge
PREA Precharge all
READ RD, RDS4, RDS8
READ A RDA, RDAS4, RDAS8
REF Refresh, fine granularity refresh
RESET Start reset procedure
SRE Self refresh entry
SRX Self refresh exit
TEN Boundary scan mode enable
WRITE WR, WRS4, WRS8 with/without CRC
WRITE A WRA, WRAS4, WRAS8 with/without CRC
ZQCL ZQ calibration long
ZQCS ZQ calibration short
Note: 1. See the Command Truth Table for more details.
Functional Description
The DDR4 SDRAM is a high-speed dynamic random-access memory internally config-
ured as sixteen banks (4 bank groups with 4 banks for each bank group) for x4/x8 devi-
ces, and as eight banks for each bank group (2 bank groups with 4 banks each) for x16
devices. The device uses double data rate (DDR) architecture to achieve high-speed op-
eration. DDR4 architecture is essentially an 8n-prefetch architecture with an interface
designed to transfer two data words per clock cycle at the I/O pins. A single read or
write access for a device module effectively consists of a single 8n-bit-wide, four-clock-
cycle-data transfer at the internal DRAM core and eight corresponding n-bit-wide, one-
half-clock-cycle data transfers at the I/O pins.
Read and write accesses to the device are burst-oriented. Accesses start at a selected lo-
cation and continue for a burst length of eight or a chopped burst of four in a program-
med sequence. Operation begins with the registration of an ACTIVE command, which is
then followed by a READ or WRITE command. The address bits registered coincident
with the ACTIVE command are used to select the bank and row to be accessed (BG[1:0]
select the bank group for x4/x8, and BG0 selects the bank group for x16; BA[1:0] select
the bank, and A[17:0] select the row. See the Addressing section for more details). The
address bits registered coincident with the READ or WRITE command are used to select
the starting column location for the burst operation, determine if the auto PRECHARGE
command is to be issued (via A10), and select BC4 or BL8 mode on-the-fly (OTF) (via
A12) if enabled in the mode register.
Prior to normal operation, the device must be powered up and initialized in a prede-
fined manner. The following sections provide detailed information covering device reset
and initialization, register definition, command descriptions, and device operation.
NOTE: The use of the NOP command is allowed only when exiting maximum power
saving mode or when entering gear-down mode.
RESET and Initialization Procedure

To ensure proper device function, the power-up and reset initialization default values
for the following mode register (MR) settings are defined as:
• Gear-down mode (MR3 A[3]): 0 = 1/2 rate
• Per-DRAM addressability (MR3 A[4]): 0 = disable
• Maximum power-saving mode (MR4 A[1]): 0 = disable
• CS to command/address latency (MR4 A[8:6]): 000 = disable
• CA parity latency mode (MR5 A[2:0]): 000 = disable
• Hard post package repair mode (MR4 A[13]): 0 = disable
• Soft post package repair mode (MR4 A[5]): 0 = disable
Power-Up and Initialization Sequence

The following sequence is required for power-up and initialization:
1. Apply power (RESET_n and TEN should be maintained below 0.2 × V DD while sup-
plies ramp up; all other inputs may be undefined). When supplies have ramped to
a valid stable level, RESET_n must be maintained below 0.2 × V DD for a minimum
of tPW_RESET_L and TEN must be maintained below 0.2 × V DD for a minimum of
700μs. CKE is pulled LOW anytime before RESET_n is de-asserted (minimum time
of 10ns). The power voltage ramp time between 300mV to V DD,min must be no
greater than 200ms, and during the ramp, V DD must be greater than or equal to
VDDQ and (VDD - V DDQ) < 0.3V. V PP must ramp at the same time or up to 10 minutes
prior to V DD, and V PP must be equal to or higher than V DD at all times. The total
time for which V PP is powered and V DD is unpowered should not exceed 360 cu-
mulative hours. After V DD has ramped and reached a stable level, RESET_n must
go high within 10 minutes. After RESET_n goes high, the initialization sequence
must be started within 3 seconds. For debug purposes, the 10 minute and 3 sec-
ond delay limits may be extended to 60 minutes each provided the DRAM is oper-
ated in this debug mode for no more than 360 cumulative hours.
During power-up, the supply slew rate is governed by the limits stated in the table
below and either condition A or condition B listed below must be met.
Table 5: Supply Power-up Slew Rate
Symbol Min Max Unit Comment

VDD_SL, VDDQ_SL, 0.004 600 V/ms Measured between 300mV and 80% of
VPP_SL supply minimum
VDD_ona N/A 200 ms VDD maximum ramp time from 300mV to
VDD minimum
VDDQ_ona N/A 200 ms VDDQ maximum ramp time from 300mV to
VDDQ minimum
Note: 1. 20 MHz band-limited measurement.
• Condition A:
– Apply V PP without any slope reversal before or at the same time as V DD and
VDDQ.
– VDD and V DDQ are driven from a single-power converter output and apply
VDD/VDDQ without any slope reversal before or at the same time as V TT and
VREFCA.
– The voltage levels on all balls other than V DD, V DDQ, V SS, and V SSQ must be less
than or equal to V DDQ and V DD on one side and must be greater than or equal
to V SSQ and V SS on the other side.
– VTT is limited to 0.76V MAX when the power ramp is complete.
– VREFCA tracks V DD/2.
• Condition B:
– Apply V PP without any slope reversal before or at the same time as V DD.
– Apply V DD without any slope reversal before or at the same time as V DDQ.
– Apply V DDQ without any slope reversal before or at the same time as V TT and
VREFCA.
– The voltage levels on all pins other than V PP, V DD, V DDQ, V SS, and V SSQ must be
less than or equal to V DDQ and V DD on one side and must be larger than or
equal to V SSQ and V SS on the other side.
2. After RESET_n is de-asserted, wait for another 500μs but no longer then 3 seconds
until CKE becomes active. During this time, the device will start internal state initi-
alization; this will be done independently of external clocks. A reasonable attempt
was made in the design to power up with the following default MR settings: gear-
down mode (MR3 A[3]): 0 = 1/2 rate; per-DRAM addressability (MR3 A[4]): 0 = dis-
able; maximum power-down (MR4 A[1]): 0 = disable; CS to command/address la-
tency (MR4 A[8:6]): 000 = disable; CA parity latency mode (MR5 A[2:0]): 000 = disa-
ble. However, it should be assumed that at power up the MR settings are unde-
fined and should be programmed as shown below.
3. Clocks (CK_t, CK_c) need to be started and stabilized for at least 10ns or 5 tCK
(whichever is larger) before CKE goes active. Because CKE is a synchronous signal,
the corresponding setup time to clock (tIS) must be met. Also, a DESELECT com-
mand must be registered (with tIS setup time to clock) at clock edge Td. After the
CKE is registered HIGH after RESET, CKE needs to be continuously registered
HIGH until the initialization sequence is finished, including expiration of tDLLK
and tZQinit.
4. The device keeps its ODT in High-Z state as long as RESET_n is asserted. Further,
the SDRAM keeps its ODT in High-Z state after RESET_n de-assertion until CKE is
registered HIGH. The ODT input signal may be in an undefined state until tIS be-
fore CKE is registered HIGH. When CKE is registered HIGH, the ODT input signal
may be statically held either LOW or HIGH. If RTT(NOM) is to be enabled in MR1,
the ODT input signal must be statically held LOW. In all cases, the ODT input sig-
nal remains static until the power-up initialization sequence is finished, including
the expiration of tDLLK and tZQinit.
5. After CKE is registered HIGH, wait a minimum of RESET CKE EXIT time, tXPR, be-
fore issuing the first MRS command to load mode register (tXPR = MAX (tXS, 5 ×
tCK).
6. Issue MRS command to load MR3 with all application settings, wait tMRD.
12. Issue MRS command to load MR0 with all application settings, wait tMOD.
13. Issue a ZQCL command to start ZQ calibration.

14. Wait for tDLLK and tZQinit to complete.
15. The device will be ready for normal operation. Once the DRAM has been initial-
ized, if the DRAM is in an idle state longer than 960ms, then either (a) REF com-
mands must be issued within tREFI constraints (specification for posting allowed)
or (b) CKE or CS_n must toggle once within every 960ms interval of idle time. For
debug purposes, the 960ms delay limit maybe extended to 60 minutes provided
the DRAM is operated in this debug mode for no more than 360 cumulative hours.
A stable valid V DD level is a set DC level (0Hz to 250 KHz) and must be no less than
VDD,min and no greater than V DD,max. If the set DC level is altered anytime after initializa-
tion, the DLL reset and calibrations must be performed again after the new set DC level
is stable. AC noise of ±60mV (greater than 250 KHz) is allowed on V DD provided the
noise doesn't alter V DD to less than V DD,min or greater than V DD,max.
A stable valid V DDQ level is a set DC level (0Hz to 250 KHz) and must be no less than
VDDQ,min and no greater than V DDQ,max. If the set DC level is altered anytime after initial-
ization, the DLL reset and calibrations must be performed again after the new set DC
level is stable. AC noise of ±60mV (greater than 250 KHz) is allowed on V DDQ provided
the noise doesn't alter V DDQ to less than V DDQ,min or greater than V DDQ,max.
A stable valid V PP level is a set DC level (0Hz to 250 KHz) and must be no less than
VPP,min and no greater than V PP,max. If the set DC level is altered anytime after initializa-
is stable. AC noise of ±120mV (greater than 250KHz) is allowed on V PP provided the
noise doesn't alter V PP to less than V PP,min or greater than V PP,max.
Figure 11: RESET and Initialization Sequence at Power-On Ramping

Ta Tb Tc Td Te Tf Tg Th Ti Tj Tk
CK_t, CK_c
tCKSRX
VPP
VDD, VDDQ
tPW_RESET_L T = 500μs
RESET_n
tIS
T (MIN) = 10ns
CKE Valid
tDLLK
tXPR tMRD tMRD tMRD tMOD tZQinit

tIS
Command Note 1 MRS MRS MRS MRS ZQCL Note 1 Valid
BG, BA MRx MRx MRx MRx Valid
tIS
tIS
ODT Static LOW in case RTT(NOM) is enabled at time Tg, otherwise static HIGH or LOW Valid
RTT
Time Break Don’t Care
Notes: 1. From time point Td until Tk, a DES command must be applied between MRS and ZQCL
commands.
2. MRS commands must be issued to all mode registers that have defined settings.
3. In general, there is no specific sequence for setting the MRS locations (except for de-
pendent or co-related features, such as ENABLE DLL in MR1 prior to RESET DLL in MR0,
for example).
4. TEN is not shown; however, it is assumed to be held LOW.
RESET Initialization with Stable Power Sequence

The following sequence is required for RESET at no power interruption initialization:
1. Assert RESET_n below 0.2 × V DD any time when reset is needed (all other inputs
may be undefined). RESET_n needs to be maintained for minimum tPW_RESET.
CKE is pulled LOW before RESET_n being de-asserted (minimum time 10ns).
2. Follow Steps 2 through 10 in the Reset and Initialization Sequence at Power-On
Ramping procedure.
When the reset sequence is complete, all counters except the refresh counters have
been reset and the device is ready for normal operation.
Figure 12: RESET Procedure at Power Stable Condition

Ta Tb Tc Td Te Tf Tg Th Ti Tj Tk
CK_t, CK_c
tCKSRX
VPP
VDD , VDDQ
tPW_RESET_S T = 500μs
RESET_n
tIS
T (MIN) = 10ns
CKE Valid
tDLLK
tXPR tMRD tMRD tMRD tMOD tZQinit

tIS
Command Note 1 MRS MRS MRS MRS ZQCL Note 1 Valid
BG, BA MRx MRx MRx MRx Valid
tIS
tIS
ODT Static LOW in case RTT(NOM) is enabled at time Tg, otherwise static HIGH or LOW Valid
RTT
Notes: 1. From time point Td until Tk, a DES command must be applied between MRS and ZQCL
commands.
2. MRS commands must be issued to all mode registers that have defined settings.
3. In general, there is no specific sequence for setting the MRS locations (except for de-
pendent or co-related features, such as ENABLE DLL in MR1 prior to RESET DLL in MR0,
for example).
4. TEN is not shown; however, it is assumed to be held LOW.
Uncontrolled Power-Down Sequence

In the event of an uncontrolled ramping down of V PP supply, V PP is allowed to be less
than V DD provided the following conditions are met:
• Condition A: V PP and V DD/VDDQ are ramping down (as part of turning off) from nor-
mal operating levels.
• Condition B: The amount that V PP may be less than V DD/VDDQ is less than or equal to
500mV.
• Condition C: The time V PP may be less than V DD is ื10ms per occurrence with a total
accumulated time in this state ื100ms.
Programming Mode Registers
• Condition D: The time V PP may be less than 2.0V and above V SS while turning off is
ื15ms per occurrence with a total accumulated time in this state ื150ms.

For application flexibility, various functions, features, and modes are programmable in
seven mode registers (MRn) provided by the device as user defined variables that must
be programmed via a MODE REGISTER SET (MRS) command. Because the default val-
ues of the mode registers are not defined, contents of mode registers must be fully ini-
tialized and/or re-initialized; that is, they must be written after power-up and/or reset
for proper operation. The contents of the mode registers can be altered by re-executing
the MRS command during normal operation. When programming the mode registers,
even if the user chooses to modify only a sub-set of the MRS fields, all address fields
within the accessed mode register must be redefined when the MRS command is is-
sued. MRS and DLL RESET commands do not affect array contents, which means these
commands can be executed any time after power-up without affecting the array con-
tents.
The MRS command cycle time, tMRD, is required to complete the WRITE operation to
the mode register and is the minimum time required between the two MRS commands
shown in the tMRD Timing figure.
Some of the mode register settings affect address/command/control input functionali-
ty. In these cases, the next MRS command can be allowed when the function being up-
dated by the current MRS command is completed. These MRS commands don’t apply
tMRD timing to the next MRS command; however, the input cases have unique MR set-
ting procedures, so refer to individual function descriptions:

• Gear-down mode
• Per-DRAM addressability
• CMD address latency
• CA parity latency mode
• VREFDQ training value
• VREFDQ training mode
• VREFDQ training range
Some mode register settings may not be supported because they are not required by
certain speed bins.
Figure 13: tMRD Timing

T0 T1 T2 Ta0 Ta1 Tb0 Tb1 Tb2 Tc0 Tc1 Tc2
CK_c
CK_t
Command Valid Valid Valid MRS2 DES DES DES DES DES MRS2 Valid
Address Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid
CKE
tMRD
Settings Old settings Updating settings
Notes: 1. This timing diagram depicts CA parity mode “disabled” case.

2. tMRD applies to all MRS commands with the following exceptions:
Gear-down mode
CA parity latency mode
CMD address latency
Per-DRAM addressability mode
VREFDQ training value, VREFDQ training mode, and VREFDQ training range
The MRS command to nonMRS command delay, tMOD, is required for the DRAM to
update features, except for those noted in note 2 in figure below where the individual
function descriptions may specify a different requirement. tMOD is the minimum time
required from an MRS command to a nonMRS command, excluding DES, as shown in
the tMOD Timing figure.
Figure 14: tMOD Timing

T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Tb0 Tb1 Tb2
CK_c
CK_t
Command Valid Valid Valid MRS2 DES DES DES DES DES Valid Valid
CKE
t
MOD
Settings Old settings Updating settings New settings
Notes: 1. This timing diagram depicts CA parity mode “disabled” case.

2. tMOD applies to all MRS commands with the following exceptions:
DLL enable, DLL RESET, Gear-down mode
VREFDQ training value, internal VREF training monitor, VREFDQ training mode, and VREFDQ
training range
Maximum power savings mode , Per-DRAM addressability mode, and CA parity latency
mode
The mode register contents can be changed using the same command and timing re-
quirements during normal operation as long as the device is in idle state; that is, all
banks are in the precharged state with tRP satisfied, all data bursts are completed, and
CKE is HIGH prior to writing into the mode register. If the RTT(NOM) feature is enabled in
the mode register prior to and/or after an MRS command, the ODT signal must contin-
uously be registered LOW, ensuring RTT is in an off state prior to the MRS command.
The ODT signal may be registered HIGH after tMOD has expired. If the RTT(NOM) feature
is disabled in the mode register prior to and after an MRS command, the ODT signal
can be registered either LOW or HIGH before, during, and after the MRS command. The
mode registers are divided into various fields depending on functionality and modes.
In some mode register setting cases, function updating takes longer than tMOD. This
type of MRS does not apply tMOD timing to the next valid command, excluding DES.
These MRS command input cases have unique MR setting procedures, so refer to indi-
vidual function descriptions.
Mode Register 0
Mode Register 0
Mode register 0 (MR0) controls various device operating modes as shown in the follow-
ing register definition table. Not all settings listed may be available on a die; only set-
tings required for speed bin support are available. MR0 is written by issuing the MRS
command while controlling the states of the BGx, BAx, and Ax address pins. The map-
ping of address pins during the MRS command is shown in the following MR0 Register
Definition table.
Table 6: Address Pin Mapping
Address BG1 BG0 BA1 BA0 A17 RAS CAS WE A13 A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Note: 1. RAS_n, CAS_n, and WE_n must be LOW during MODE REGISTER SET command.
Table 7: MR0 Register Definition
Mode
Register Description
21 RFU
0 = Must be programmed to 0
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
17 N/A on 4Gb and 8Gb, RFU
1 = Reserved
13,11:9 WR (WRITE recovery)/RTP (READ-to-PRECHARGE)
0000 = 10 / 5 clocks1
0001 = 12 / 6 clocks
0010 = 14 / 7 clocks1
0011 = 16 / 8 / clocks
0100 = 18 / 9 clocks1
0101 = 20 /10 clocks
0110 = 24 / 12 clocks
0111 = 22 / 11 clocks1
1000 = 26 / 13 clocks1
1001 = 28 / 14 clocks2
1010 through 1111 = Reserved
Mode Register 0
Table 7: MR0 Register Definition (Continued)
Mode
8 DLL reset
0 = No
1 = Yes
7 Test mode (TM) – Manufacturer use only
0 = Normal operating mode, must be programmed to 0
12, 6:4, 2 CAS latency (CL) – Delay in clock cycles from the internal READ command to first data-out
00000 = 9 clocks1
00001 = 10 clocks
00010 = 11 clocks1
00011 = 12 clocks
00100 = 13 clocks1
00101 = 14 clocks
00110 = 15 clocks1
00111 = 16 clocks
01000 = 18 clocks
01001 = 20 clocks
01010 = 22 clocks
01011 = 24 clocks
01100 = 23 clocks1
01101 = 17 clocks1
01110 = 19 clocks1
01111 = 21 clocks 1
10000 = 25 clocks
10001 = 26 clocks
10011 = 28 clocks
10100 = 29 clocks1
10101 = 30 clocks
10110 = 31 clocks1
10111 = 32 clocks
3 Burst type (BT) – Data burst ordering within a READ or WRITE burst access
0 = Nibble sequential
1 = Interleave
1:0 Burst length (BL) – Data burst size associated with each read or write access
00 = BL8 (fixed)
01 = BC4 or BL8 (on-the-fly)
10 = BC4 (fixed)
11 = Reserved
Notes: 1. Not allowed when 1/4 rate gear-down mode is enabled.

2. If WR requirement exceeds 28 clocks or RTP exceeds 14 clocks, WR should be set to 28
clocks and RTP should be set to 14 clocks.
Burst Length, Type, and Order

Accesses within a given burst may be programmed to sequential or interleaved order.
The ordering of accesses within a burst is determined by the burst length, burst type,
Mode Register 0
and the starting column address as shown in the following table. Burst length options
include fixed BC4, fixed BL8, and on-the-fly (OTF), which allows BC4 or BL8 to be selec-
ted coincidentally with the registration of a READ or WRITE command via A12/BC_n.
Table 8: Burst Type and Burst Order

Note 1 applies to the entire table
Starting
Burst READ/ Column Address Burst Type = Sequential Burst Type = Interleaved
Length WRITE (A[2, 1, 0]) (Decimal) (Decimal) Notes
BC4 READ 000 0, 1, 2, 3, T, T, T, T 0, 1, 2, 3, T, T, T, T 2, 3
001 1, 2, 3, 0, T, T, T, T 1, 0, 3, 2, T, T, T, T 2, 3
010 2, 3, 0, 1, T, T, T, T 2, 3, 0, 1, T, T, T, T 2, 3
011 3, 0, 1, 2, T, T, T, T 3, 2, 1, 0, T, T, T, T 2, 3
100 4, 5, 6, 7, T, T, T, T 4, 5, 6, 7, T, T, T, T 2, 3
101 5, 6, 7, 4, T, T, T, T 5, 4, 7, 6, T, T, T, T 2, 3
110 6, 7, 4, 5, T, T, T, T 6, 7, 4, 5, T, T, T, T 2, 3
111 7, 4, 5, 6, T, T, T, T 7, 6, 5, 4, T, T, T, T 2, 3
WRITE 0, V, V 0, 1, 2, 3, X, X, X, X 0, 1, 2, 3, X, X, X, X 2, 3
1, V, V 4, 5, 6, 7, X, X, X, X 4, 5, 6, 7, X, X, X, X 2, 3
BL8 READ 000 0, 1, 2, 3, 4, 5, 6, 7 0, 1, 2, 3, 4, 5, 6, 7
001 1, 2, 3, 0, 5, 6, 7, 4 1, 0, 3, 2, 5, 4, 7, 6
010 2, 3, 0, 1, 6, 7, 4, 5 2, 3, 0, 1, 6, 7, 4, 5
011 3, 0, 1, 2, 7, 4, 5, 6 3, 2, 1, 0, 7, 6, 5, 4
100 4, 5, 6, 7, 0, 1, 2, 3 4, 5, 6, 7, 0, 1, 2, 3
101 5, 6, 7, 4, 1, 2, 3, 0 5, 4, 7, 6, 1, 0, 3, 2
110 6, 7, 4, 5, 2, 3, 0, 1 6, 7, 4, 5, 2, 3, 0, 1
111 7, 4, 5, 6, 3, 0, 1, 2 7, 6, 5, 4, 3, 2, 1, 0
WRITE V, V, V 0, 1, 2, 3, 4, 5, 6, 7 0, 1, 2, 3, 4, 5, 6, 7 3
Notes: 1. 0...7 bit number is the value of CA[2:0] that causes this bit to be the first read during a
burst.
2. When setting burst length to BC4 (fixed) in MR0, the internal WRITE operation starts
two clock cycles earlier than for the BL8 mode, meaning the starting point for tWR and
tWTR will be pulled in by two clocks. When setting burst length to OTF in MR0, the in-
ternal WRITE operation starts at the same time as a BL8 (even if BC4 was selected during
column time using A12/BC4_n) meaning that if the OTF MR0 setting is used, the starting
point for tWR and tWTR will not be pulled in by two clocks as described in the BC4
(fixed) case.
3. T = Output driver for data and strobes are in High-Z.
V = Valid logic level (0 or 1), but respective buffer input ignores level on input pins.
X = “Don’t Care.”
CAS Latency
The CAS latency (CL) setting is defined in the MR0 Register Definition table. CAS laten-
cy is the delay, in clock cycles, between the internal READ command and the availability
of the first bit of output data. The device does not support half-clock latencies. The
Mode Register 0
overall read latency (RL) is defined as additive latency (AL) + CAS latency (CL): RL = AL +
CL.
Test Mode
The normal operating mode is selected by MR0[7] and all other bits set to the desired
values shown in the MR0 Register Definition table. Programming MR0[7] to a value of 1
places the device into a DRAM manufacturer-defined test mode to be used only by the
manufacturer, not by the end user. No operations or functionality is specified if MR0[7]
= 1.
Write Recovery (WR)/READ-to-PRECHARGE

The programmed write recovery (WR) value is used for the auto precharge feature along
with tRP to determine tDAL. WR for auto precharge (MIN) in clock cycles is calculated
by dividing tWR (in ns) by tCK (in ns) and rounding to the next integer using the round-
ing algorithms found in the Converting Time-Based Specifications to Clock-Based Re-
quirements section. The WR value must be programmed to be equal to or larger than
tWR (MIN). When both DM and write CRC are enabled in the mode register, the device
calculates CRC before sending the write data into the array; tWR values will change
when enabled. If there is a CRC error, the device blocks the WRITE operation and dis-
cards the data.
Internal READ-to-PRECHARGE (RTP) command delay for auto precharge (MIN) in
clock cycles is calculated by dividing tRTP (in ns) by tCK (in ns) and rounding to the next
integer using the rounding algorithms found in the Converting Time-Based Specifica-
tions to Clock-Based Requirements section. The RTP value in the mode register must be
programmed to be equal to or larger than RTP (MIN). The programmed RTP value is
used with tRP to determine the ACT timing to the same bank.
DLL RESET
The DLL reset bit is self-clearing, meaning that it returns to the value of 0 after the DLL
RESET function has been issued. After the DLL is enabled, a subsequent DLL RESET
should be applied. Any time the DLL RESET function is used, tDLLK must be met before
functions requiring the DLL can be used. Such as READ commands or synchronous
ODT operations, for example.
Mode Register 1
Mode Register 1
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
1 = Reserved
12 Data output disable (Qoff) – Output buffer disable
0 = Enabled (normal operation)
1 = Disabled (both ODI and RTT)
11 Termination data strobe (TDQS) – Additional termination pins (x8 configuration only)
0 = TDQS disabled
1 = TDQS enabled
Mode Register 1
Mode
10, 9, 8 Nominal ODT (RTT(NOM) – Data bus termination setting
000 = RTT(NOM) disabled
001 = RZQ/4 (60 ohm)
010 = RZQ/2 (120 ohm)
011 = RZQ/6 (40 ohm)
100 = RZQ/1 (240 ohm)
101 = RZQ/5 (48 ohm)
110 = RZQ/3 (80 ohm)
111 = RZQ/7 (34 ohm)
7 Write leveling (WL) – Write leveling mode
0 = Disabled (normal operation)
1 = Enabled (enter WL mode)
13, 6, 5 Rx CTLE Control
000 = Vendor Default
001 = Vendor Defined
4, 3 Additive latency (AL) – Command additive latency setting
00 = 0 (AL disabled)
01 = CL - 11
10 = CL - 2
11 = Reserved
2, 1 Output driver impedance (ODI) – Output driver impedance setting
00 = RZQ/7 (34 ohm)
01 = RZQ/5 (48 ohm)
10 = Reserved (Although not JEDEC-defined and not tested, this setting will provide RZQ/6 or 40 ohm)
11 = Reserved
0 DLL enable – DLL enable feature
0 = DLL disabled
1 = DLL enabled (normal operation)
Note: 1. Not allowed when 1/4 rate gear-down mode is enabled.
DLL Enable/DLL Disable

The DLL must be enabled for normal operation and is required during power-up initial-
ization and upon returning to normal operation after having the DLL disabled. During
normal operation (DLL enabled with MR1[0]) the DLL is automatically disabled when
entering the SELF REFRESH operation and is automatically re-enabled upon exit of the
SELF REFRESH operation. Any time the DLL is enabled and subsequently reset, tDLLK
clock cycles must occur before a READ or SYNCHRONOUS ODT command can be is-
sued to allow time for the internal clock to be synchronized with the external clock. Fail-
Mode Register 1
ing to wait for synchronization to occur may result in a violation of the tDQSCK, tAON,
or tAOF parameters.
During tDLLK, CKE must continuously be registered HIGH. The device does not require
DLL for any WRITE operation, except when R TT(WR) is enabled and the DLL is required
for proper ODT operation.
The direct ODT feature is not supported during DLL off mode. The ODT resistors must
be disabled by continuously registering the ODT pin LOW and/or by programming the
RTT(NOM) bits MR1[9,6,2] = 000 via an MRS command during DLL off mode.
The dynamic ODT feature is not supported in DLL off mode; to disable dynamic ODT
externally, use the MRS command to set RTT(WR), MR2[10:9] = 00.
Output Driver Impedance Control

The output driver impedance of the device is selected by MR1[2,1], as shown in the MR1
Register Definition table.
ODT RTT(NOM) Values

The device is capable of providing three different termination values: RTT(Park), RTT(NOM),
and RTT(WR). The nominal termination value, R TT(NOM), is programmed in MR1. A sepa-
rate value, RTT(WR), may be programmed in MR2 to enable a unique RTT value when
ODT is enabled during WRITE operations. The R TT(WR) value can be applied during
WRITE commands even when RTT(NOM) is disabled. A third RTT value, RTT(Park), is pro-
gramed in MR5. RTT(Park) provides a termination value when the ODT signal is LOW.
Additive Latency
The ADDITIVE LATENCY (AL) operation is supported to make command and data
buses efficient for sustainable bandwidths in the device. In this operation, the device al-
lows a READ or WRITE command (either with or without auto precharge) to be issued
immediately after the ACTIVATE command. The command is held for the time of AL be-
fore it is issued inside the device. READ latency (RL) is controlled by the sum of the AL
and CAS latency (CL) register settings. WRITE latency (WL) is controlled by the sum of
the AL and CAS WRITE latency (CWL) register settings.
Table 11: Additive Latency (AL) Settings
A4 A3 AL
0 0 0 (AL disabled)
0 1 CL - 1
1 0 CL - 2
1 1 Reserved
Note: 1. AL has a value of CL - 1 or CL - 2 based on the CL values programmed in the MR0 regis-
ter.
Rx CTLE Control
The Mode Register for Rx CTLE Control MR1[A13,A6,A5] is vendor specific. Since CTLE
circuits can not be typically bypassed a disable option is not provided. Instead, a vendor
optimized setting is given. It should be noted that the settings are not specifically linear
Mode Register 1
in relationship to the vendor optimized setting, so the host may opt to instead walk
through all the provided options and use the setting that works best in their environ-
ment.
Write Leveling
For better signal integrity, the device uses fly-by topology for the commands, addresses,
control signals, and clocks. Fly-by topology benefits from a reduced number of stubs
and their lengths, but it causes flight-time skew between clock and strobe at every
DRAM on the DIMM. This makes it difficult for the controller to maintain tDQSS, tDSS,
and tDSH specifications. Therefore, the device supports a write leveling feature that al-
lows the controller to compensate for skew.
Output Disable
The device outputs may be enabled/disabled by MR1[12] as shown in the MR1 Register
Definition table. When MR1[12] is enabled (MR1[12] = 1) all output pins (such as DQ
and DQS) are disconnected from the device, which removes any loading of the output
drivers. For example, this feature may be useful when measuring module power. For
normal operation, set MR1[12] to 0.
Termination Data Strobe

Termination data strobe (TDQS) is a feature of the x8 device and provides additional
termination resistance outputs that may be useful in some system configurations. Be-
cause this function is available only in a x8 configuration, it must be disabled for x4 and
x16 configurations.
While TDQS is not supported in x4 or x16 configurations, the same termination resist-
ance function that is applied to the TDQS pins is applied to the DQS pins when enabled
via the mode register.
The TDQS, DBI, and DATA MASK (DM) functions share the same pin. When the TDQS
function is enabled via the mode register, the DM and DBI functions are not supported.
When the TDQS function is disabled, the DM and DBI functions can be enabled sepa-
rately.
Table 12: TDQS Function Matrix
TDQS Data Mask (DM) WRITE DBI READ DBI

Disabled Enabled Disabled Enabled or disabled
Disabled Enabled Enabled or disabled
Disabled Disabled Enabled or disabled
Enabled Disabled Disabled Disabled
Mode Register 2
Mode Register 2
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
1 = Reserved
13 RFU
1 = Reserved
12 WRITE data bus CRC
0 = Disabled
1 = Enabled
Mode Register 2
Mode
11:9 Dynamic ODT (RTT(WR)) – Data bus termination setting during WRITEs
000 = RTT(WR) disabled (WRITE does not affect RTT value)
001 = RZQ/2 (120 ohm)
010 = RZQ/1 (240 ohm)
011 = High-Z
100 = RZQ/3 (80 ohm)
101 = Reserved
110 = Reserved
111 = Reserved
7:6 Low-power auto self refresh (LPASR) – Mode summary
00 = Manual mode - Normal operating temperature range (TC: –40°C–85°C)
01 = Manual mode - Reduced operating temperature range (TC: –40°C–45°C)
10 = Manual mode - Extended operating temperature range (TC: –40°C–125°C)
11 = ASR mode - Automatically switching among all modes
5:3 CAS WRITE latency (CWL) – Delay in clock cycles from the internal WRITE command to first data-in
1tCK WRITE preamble
000 = 9 (DDR4-1600)1
001 = 10 (DDR4-1866)
010 = 11 (DDR4-2133/1600)1
011 = 12 (DDR4-2400/1866)
100 = 14 (DDR4-2666/2133)
101 = 16 (DDR4-2933,3200/2400)
110 = 18 (DDR4-2666)
111 = 20 (DDR4-2933, 3200)
CAS WRITE latency (CWL) – Delay in clock cycles from the internal WRITE command to first data-in
2tCK WRITE preamble
000 = N/A
001 = N/A
010 = N/A
011 = N/A
100 = 14 (DDR4-2400)
101 = 16 (DDR4-2666/2400)
110 = 18 (DDR4-2933, 3200/2666)
111 = 20 (DDR4-2933, 3200)
8, 2 RFU
1 = Reserved
1:0 RFU
1 = Reserved
Mode Register 2
CAS WRITE Latency

CAS WRITE latency (CWL) is defined by MR2[5:3] as shown in the MR2 Register Defini-
tion table. CWL is the delay, in clock cycles, between the internal WRITE command and
the availability of the first bit of input data. The device does not support any half-clock
latencies. The overall WRITE latency (WL) is defined as additive latency (AL) + parity la-
tency (PL) + CAS WRITE latency (CWL): WL = AL +PL + CWL.
Low-Power Auto Self Refresh

Low-power auto self refresh (LPASR) is supported in the device. Applications requiring
SELF REFRESH operation over different temperature ranges can use this feature to opti-
mize the IDD6 current for a given temperature range as specified in the MR2 Register
Definition table.
Dynamic ODT
In certain applications and to further enhance signal integrity on the data bus, it is de-
sirable to change the termination strength of the device without issuing an MRS com-
mand. This may be done by configuring the dynamic ODT (R TT(WR)) settings in
MR2[11:9]. In write leveling mode, only RTT(NOM) is available.
Write Cyclic Redundancy Check Data Bus

The write cyclic redundancy check (CRC) data bus feature during writes has been added
to the device. When enabled via the mode register, the data transfer size goes from the
normal 8-bit (BL8) frame to a larger 10-bit UI frame, and the extra two UIs are used for
the CRC information.
Mode Register 3
Mode Register 3
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
1 = Reserved
13 RFU
1 = Reserved
12:11 Multipurpose register (MPR) – Read format
00 = Serial
01 = Parallel
10 = Staggered
11 = Reserved
10:9 WRITE CMD latency when CRC/DM enabled
00 = 4CK (DDR4-1600)
01 = 5CK (DDR4-1866/2133/2400/2666)
10 = 6CK (DDR4-2933/3200)
11 = Reserved
Mode Register 3
Mode
8:6 Fine granularity refresh mode
000 = Normal mode (fixed 1x)
001 = Fixed 2x
010 = Fixed 4x
011 = Reserved
100 = Reserved
101 = On-the-fly 1x/2x
110 = On-the-fly 1x/4x
111 = Reserved
5 Temperature sensor status
0 = Disabled
1 = Enabled
4 Per-DRAM addressability
0 = Normal operation (disabled)
1 = Enable
3 Gear-down mode – Ratio of internal clock to external data rate
0 = [1:1]; (1/2 rate data)
1 = [2:1]; (1/4 rate data)
2 Multipurpose register (MPR) access
0 = Normal operation
1 = Data flow from MPR
1:0 MPR page select
00 = Page 0
01 = Page 1
10 = Page 2
11 = Page 3 (restricted for DRAM manufacturer use only)
Multipurpose Register
The multipurpose register (MPR) is used for several features:
• Readout of the contents of the MRn registers
• WRITE and READ system patterns used for data bus calibration
• Readout of the error frame when the command address parity feature is enabled
To enable MPR, issue an MRS command to MR3[2] = 1. MR3[12:11] define the format of
read data from the MPR. Prior to issuing the MRS command, all banks must be in the
idle state (all banks precharged and tRP met). After MPR is enabled, any subsequent RD
or RDA commands will be redirected to a specific mode register.
The mode register location is specified with the READ command using address bits. The
MR is split into upper and lower halves to align with a burst length limitation of 8. Pow-
er-down mode, SELF REFRESH, and any other nonRD/RDA or nonWR/WRA com-
mands are not allowed during MPR mode. The RESET function is supported during
MPR mode, which requires device re-initialization.
Mode Register 3
WRITE Command Latency When CRC/DM is Enabled

The WRITE command latency (WCL) must be set when both write CRC and DM are en-
abled for write CRC persistent mode. This provides the extra time required when com-
pleting a WRITE burst when write CRC and DM are enabled. This means at data rates
less than or equal to 1600 MT/s then 4nCK is used, 5nCK or 6nCK are not allowed; at
data rates greater than 1600 MT/s and less than or equal to 2666 MT/s then 5nCK is
used, 4nCK or 6nCK are not allowed; and at data rates greater than 2666 MT/s and less
than or equal to 3200 MT/s then 6nCK is used; 4nCK or 5nCK are not allowed.
Fine Granularity Refresh Mode

This mode had been added to DDR4 to help combat the performance penalty due to
refresh lockout at high densities. Shortening tRFC and decreasing cycle time allows
more accesses to the chip and allows for increased scheduling flexibility.
Temperature Sensor Status

This mode directs the DRAM to update the temperature sensor status at MPR Page 2,
MPR0 [4,3]. The temperature sensor setting should be updated within 32ms; when an
MPR read of the temperature sensor status bits occurs, the temperature sensor status
should be no older than 32ms.
Per-DRAM Addressability
This mode allows commands to be masked on a per device basis providing any device
in a rank (devices sharing the same command and address signals) to be programmed
individually. As an example, this feature can be used to program different ODT or V REF
values on DRAM devices within a given rank.
Gear-Down Mode
The device defaults in 1/2 rate (1N) clock mode and uses a low frequency MRS com-
mand followed by a sync pulse to align the proper clock edge for operating the control
lines CS_n, CKE, and ODT when in 1/4 rate (2N) mode. For operation in 1/2 rate mode,
no MRS command or sync pulse is required.
Mode Register 4
Mode Register 4
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Note: 1. RAS_n, CAS_n, and WE_n must be LOW during MODE REGISTER SET (MRS) command.
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
1 = Reserved
13 Hard Post Package Repair (hPPR mode)
0 = Disabled
1 = Enabled
12 WRITE preamble setting
0 = 1tCK toggle1
1 = 2tCK toggle (When operating in 2tCK WRITE preamble mode, CWL must be programmed to a value at
least 1 clock greater than the lowest CWL setting supported in the applicable tCK range.)
11 READ preamble setting
0 = 1tCK toggle1
1 = 2tCK toggle
10 READ preamble training
0 = Disabled
1 = Enabled
Mode Register 4
Mode
9 Self refresh abort mode
0 = Disabled
1 = Enabled
8:6 CMD (CAL) address latency
000 = 0 clocks (disabled)
001 =3 clocks1
010 = 4 clocks
011 = 5 clocks1
100 = 6 clocks
101 = 8 clocks
110 = Reserved
111 = Reserved
5 soft Post Package Repair (sPPR mode)
0 = Disabled
1 = Enabled
4 Internal VREF monitor
0 = Disabled
1 = Enabled
3 Temperature controlled refresh mode
0 = Disabled
1 = Enabled
2 Temperature controlled refresh range
0 = Normal temperature mode
1 = Extended temperature mode
1 Maximum power savings mode
0 = Normal operation
1 = Enabled
0 RFU
1 = Reserved
Hard Post Package Repair Mode

The hard post package repair (hPPR) mode feature is JEDEC optional for 4Gb DDR4
memories. Performing an MPR read to page 2 MPR0 [7] indicates whether hPPR mode is
available (A7 = 1) or not available (A7 = 0). hPPR mode provides a simple and easy repair
method of the device after placed in the system. One row per bank can be repaired. The
repair process is irrevocable so great care should be exercised when using.
Soft Post Package Repair Mode

The soft post package repair (sPPR) mode feature is JEDEC optional for 4Gb and 8Gb
DDR4 memories. Performing an MPR read to page 2 MPR0 [6] indicates whether sPPR
mode is available (A6 = 1) or not available (A6 = 0). sPPR mode provides a simple and
Mode Register 4
easy repair method of the device after placed in the system. One row per bank can be
repaired. The repair process is revocable by either doing a reset or power-down or by
rewriting a new address in the same bank.
WRITE Preamble
Programmable WRITE preamble, tWPRE, can be set to 1tCK or 2tCK via the MR4 register.
The 1tCK setting is similar to DDR3. However, when operating in 2tCK WRITE preamble
mode, CWL must be programmed to a value at least 1 clock greater than the lowest CWL
setting supported in the applicable tCK range.
When operating in 2tCK WRITE preamble mode, CWL must be programmed to a value
at least 1 clock greater than the lowest CWL setting supported in the applicable tCK
range. Some even settings will require addition of 2 clocks. If the alternate longer CWL
was used, the additional clocks will not be required.
READ Preamble
Programmable READ preamble tRPRE can be set to 1tCK or 2tCK via the MR4 register.
Both the 1tCK and 2tCK DDR4 preamble settings are different from that defined for the
DDR3 SDRAM. Both DDR4 READ preamble settings may require the memory controller
to train (or read level) its data strobe receivers using the READ preamble training.
READ Preamble Training

Programmable READ preamble training can be set to 1tCK or 2tCK. This mode can be
used by the memory controller to train or READ level its data strobe receivers.
Temperature-Controlled Refresh
When temperature-controlled refresh mode is enabled, the device may adjust the inter-
nal refresh period to be longer than tREFI of the normal temperature range by skipping
external REFRESH commands with the proper gear ratio. For example, the DRAM tem-
perature sensor detected less than 45°C. Normal temperature mode covers the range of
–40°C to 85°C, while the extended temperature range covers –40°C to 125°C.
Command Address Latency

COMMAND ADDRESS LATENCY (CAL) is a power savings feature and can be enabled
or disabled via the MRS setting. CAL is defined as the delay in clock cycles (tCAL) be-
tween a CS_n registered LOW and its corresponding registered command and address.
The value of CAL (in clocks) must be programmed into the mode register and is based
on the roundup (in clocks) of [tCK(ns)/tCAL(ns)].
Internal VREF Monitor

The device generates its own internal V REFDQ. This mode may be enabled during V REFDQ
training, and when enabled, V REF,time-short and V REF,time-long need to be increased by 10ns
if DQ0, DQ1, DQ2, or DQ3 have 0pF loading. An additional 15ns per pF of loading is also
needed.
Mode Register 4
Maximum Power Savings Mode

This mode provides the lowest power mode where data retention is not required. When
the device is in the maximum power saving mode, it does not need to guarantee data
retention or respond to any external command (except the MAXIMUM POWER SAVING
MODE EXIT command and during the assertion of RESET_n signal LOW).
Mode Register 5
Mode Register 5
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
1 = Reserved
13 RFU
1 = Reserved
12 Data bus inversion (DBI) – READ DBI enable
0 = Disabled
1 = Enabled
11 Data bus inversion (DBI) – WRITE DBI enable
0 = Disabled
1 = Enabled
10 Data mask (DM)
0 = Disabled
1 = Enabled
Mode Register 5
Mode
9 CA parity persistent error mode
0 = Disabled
1 = Enabled
8:6 Parked ODT value (RTT(Park))
000 = RTT(Park) disabled
001 = RZQ/4 (60 ohm)
010 = RZQ/2 (120 ohm)
011 = RZQ/6 (40 ohm)
100 = RZQ/1 (240 ohm)
101 = RZQ/5 (48 ohm)
110 = RZQ/3 (80 ohm)
111 = RZQ/7 (34 ohm)
5 ODT input buffer for power-down
0 = Buffer enabled
1 = Buffer disabled
4 CA parity error status
0 = Clear
1 = Error
3 CRC error status
0 = Clear
1 = Error
2:0 CA parity latency mode
000 = Disable
001 = 4 clocks (DDR4-1600/1866/2133)
010 = 5 clocks (DDR4-2400/2666)1
011 = 6 clocks (DDR4-2933/3200)
100 = Reserved
101 = Reserved
110 = Reserved
111 = Reserved
Data Bus Inversion

The DATA BUS INVERSION (DBI) function has been added to the device and is suppor-
ted only for x8 and x16 configurations (x4 is not supported). The DBI function shares a
common pin with the DM and TDQS functions. The DBI function applies to both READ
and WRITE operations; Write DBI cannot be enabled at the same time the DM function
is enabled. Refer to the TDQS Function Matrix table for valid configurations for all three
functions (TDQS/DM/DBI). DBI is not allowed during MPR READ operation; during an
MPR read, the DRAM ignores the read DBI enable setting in MR5 bit A12.
DBI is not supported for 3DS devices and should be disabled in MR5.
Mode Register 5
Data Mask
The DATA MASK (DM) function, also described as a partial write, has been added to the
device and is supported only for x8 and x16 configurations (x4 is not supported). The
DM function shares a common pin with the DBI and TDQS functions. The DM function
applies only to WRITE operations and cannot be enabled at the same time the write DBI
function is enabled. Refer to the TDQS Function Matrix table for valid configurations for
all three functions (TDQS/DM/DBI).
CA Parity Persistent Error Mode

Normal CA parity mode (CA parity persistent mode disabled) no longer performs CA
parity checking while the parity error status bit remains set at 1. However, with CA pari-
ty persistent mode enabled, CA parity checking continues to be performed when the
parity error status bit is set to a 1.
ODT Input Buffer for Power-Down

This feature determines whether the ODT input buffer is on or off during power-down.
If the input buffer is configured to be on (enabled during power-down), the ODT input
signal must be at a valid logic level. If the input buffer is configured to be off (disabled
during power-down), the ODT input signal may be floating and the device does not pro-
vide RTT(NOM) termination. However, the device may provide RTT(Park) termination de-
pending on the MR settings. This is primarily for additional power savings.
CA Parity Error Status

The device will set the error status bit to 1 upon detecting a parity error. The parity error
status bit remains set at 1 until the device controller clears it explicitly using an MRS
command.
CRC Error Status

The device will set the error status bit to 1 upon detecting a CRC error. The CRC error
status bit remains set at 1 until the device controller clears it explicitly using an MRS
command.
CA Parity Latency Mode

CA parity is enabled when a latency value, dependent on tCK, is programmed; this ac-
counts for parity calculation delay internal to the device. The normal state of CA parity
is to be disabled. If CA parity is enabled, the device must ensure there are no parity er-
rors before executing the command. CA parity signal (PAR) covers ACT_n, RAS_n/A16 ,
CAS_n/A15, WE_n/A14, and the address bus including bank address and bank group
bits. The control signals CKE, ODT, and CS_n are not included in the parity calculation.
Mode Register 6
Mode Register 6
Definition table.
bus _n _n _n
Mode 21 20 19 18 17 – – – 13 12 11 10 9 8 7 6 5 4 3 2 1 0
register
Mode
21 RFU
1 = Reserved
20:18 MR select
000 = MR0
001 = MR1
010 = MR2
011 = MR3
100 = MR4
101 = MR5
110 = MR6
111 = DNU
17 NA on 4Gb and 8Gb, RFU
1 = Reserved
12:10 tCCD_L
000 = 4 clocks (Data rateื 1333 Mb/s)

001 = 5 clocks (1333 Mb/s <Data rate ื1866 Mb/s)
101 = Reserved
110 = Reserved
111 = Reserved
Mode Register 6
Mode
13, 9, 8 RFU
Default = 000; Must be programmed to 000
001 = Reserved
010 = Reserved
011 = Reserved
100 = Reserved
101 = Reserved
110 = Reserved
111 = Reserved
7 VREF Calibration Enable
0 = Disable
1 = Enable
6 VREF Calibration Range
0 = Range 1
1 = Range 2
5:0 VREF Calibration Value
See the VREFDQ Range and Levels table in the VREFDQ Calibration section
tCCD_L Programming
The device controller must program the correct tCCD_L value. tCCD_L will be program-
med according to the value defined per operating frequency in the AC parameter table.
Although JEDEC specifies the larger of 5nCK or Xns, Micron's DRAM supports the larger
of 4nCK or Xns.
VREFDQ Calibration Enable

VREFDQ calibration is where the device internally generates its own V REFDQ to be used by
the DQ input receivers. The V REFDQ value will be output on any DQ of DQ[3:0] for evalu-
ation only. The device controller is responsible for setting and calibrating the internal
VREFDQ level using an MRS protocol (adjust up, adjust down, and so on). It is assumed
that the controller will use a series of writes and reads in conjunction with V REFDQ ad-
justments to optimize and verify the data eye. Enabling V REFDQ calibration must be used
whenever values are being written to the MR6[6:0] register.
VREFDQ Calibration Range

The device defines two V REFDQ calibration ranges: Range 1 and Range 2. Range 1 sup-
ports V REFDQ between 60% and 92% of V DDQ while Range 2 supports V REFDQ between
45% and 77% of V DDQ, as seen in V REFDQ Specification table. Although not a restriction,
Range 1 was targeted for module-based designs and Range 2 was added to target point-
to-point designs.
Mode Register 6
VREFDQ Calibration Value

Fifty settings provide approximately 0.65% of granularity steps sizes for both Range 1
and Range 2 of V REFDQ, as seen in V REFDQ Range and Levels table in the V REFDQ Calibra-
tion section.
4gb_auto_ddr4_sdram_z90b_z10B.pdf - Rev. I 05/19 EN
CCMTD-1725822587-10418
Truth Tables
Table 23: Truth Table – Command

Notes 1–5 apply to the entire table; Note 6 applies to all READ/WRITE commands
RAS_n/A16
CAS_n/A15
WE_n/A14
A12/BC_n
A[13,11]
BA [1:0]
Symbol
BG[1:0]
A10/AP
ACT_n
A[9:0]
C[2:0]
CS_n
Prev. Pres.
Function CKE CKE Notes
MODE REGISTER SET MRS H H L H L L L BG BA V OP code 7
REFRESH REF H H L H L L H V V V V V V V
Self refresh entry SRE H L L H L L H V V V V V V V 8, 9, 10
Self refresh exit SRX L H H X X X X X X X X X X X 8, 9, 10,
L H H H H V V V V V V V 11
Single-bank PRECHARGE PRE H H L H L H L BG BA V V V L V
PRECHARGE all banks PREA H H L H L H L V V V V V H V
Reserved for future use RFU H H L H L H H RFU
Bank ACTIVATE ACT H H L L Row address (RA) BG BA V Row address (RA)
WRITE BL8 fixed, BC4 fixed WR H H L H H L L BG BA V V V L CA
67
BC4OTF WRS4 H H L H H L L BG BA V L V L CA

BL8OTF WRS8 H H L H H L L BG BA V H V L CA
WRITE BL8 fixed, BC4 fixed WRA H H L H H L L BG BA V V V H CA
Micron Technology, Inc. reserves the right to change products or specifications without notice.
with auto BC4OTF WRAS4 H H L H H L L BG BA V L V H CA

precharge
BL8OTF WRAS8 H H L H H L L BG BA V H V H CA
READ BL8 fixed, BC4 fixed RD H H L H H L H BG BA V V V L CA
BC4OTF RDS4 H H L H H L H BG BA V L V L CA
BL8OTF RDS8 H H L H H L H BG BA V H V L CA
READ BL8 fixed, BC4 fixed RDA H H L H H L H BG BA V V V H CA
with auto BC4OTF RDAS4 H H L H H L H BG BA V L V H CA
2016 Micron Technology, Inc. All rights reserved.
precharge
BL8OTF RDAS8 H H L H H L H BG BA V H V H CA
NO OPERATION NOP H H L H H H H V V V V V V V 12
Device DESELECTED DES H H H X X X X X X X X X X X 13
Truth Tables
Power-down entry PDE H L H X X X X X X X X X X X 10, 14
Power-down exit PDX L H H X X X X X X X X X X X 10, 14
ZQ CALIBRATION LONG ZQCL H H L H H H L X X X X X H X
ZQ CALIBRATION SHORT ZQCS H H L H H H L X X X X X L X
Truth Tables
Notes: 1. • BG = Bank group address

• BA = Bank address
• RA = Row address
• CA = Column address
• BC_n = Burst chop
• X = “Don’t Care”
• V = Valid
2. All DDR4 SDRAM commands are defined by states of CS_n, ACT_n, RAS_n/A16, CAS_n/
A15, WE_n/A14, and CKE at the rising edge of the clock. The MSB of BG, BA, RA, and CA
are device density- and configuration-dependent. When ACT_n = H, pins RAS_n/A16,
CAS_n/A15, and WE_n/A14 are used as command pins RAS_n, CAS_n, and WE_n, respec-
tively. When ACT_n = L, pins RAS_n/A16, CAS_n/A15, and WE_n/A14 are used as address
pins A16, A15, and A14, respectively.
3. RESET_n is enabled LOW and is used only for asynchronous reset and must be main-
tained HIGH during any function.
4. Bank group addresses (BG) and bank addresses (BA) determine which bank within a
bank group is being operated upon. For MRS commands, the BG and BA selects the spe-
cific mode register location.
5. V means HIGH or LOW (but a defined logic level), and X means either defined or unde-
fined (such as floating) logic level.
6. READ or WRITE bursts cannot be terminated or interrupted, and fixed/on-the-fly (OTF)
BL will be defined by MRS.
7. During an MRS command, A17 is RFU and is device density- and configuration-depend-
ent.
8. The state of ODT does not affect the states described in this table. The ODT function is
not available during self refresh.
9. VPP and VREF (VREFCA) must be maintained during SELF REFRESH operation.
10. Refer to the Truth Table – CKE table for more details about CKE transition.
11. Controller guarantees self refresh exit to be synchronous. DRAM implementation has
the choice of either synchronous or asynchronous.
12. The NO OPERATION (NOP) command may be used only when exiting maximum power
saving mode or when entering gear-down mode.
13. The NOP command may not be used in place of the DESELECT command.
14. The power-down mode does not perform any REFRESH operation.
Truth Tables
Table 24: Truth Table – CKE

Notes 1–7, 9, and 20 apply to the entire table
CKE
Previous Cycle Present Cycle
Current State (n - 1) (n) Command (n) Action (n) Notes
Power-down L L X Maintain power-down 8, 10, 11
L H DES Power-down exit 8, 10, 12
Self refresh L L X Maintain self refresh 11, 13
L H DES Self refresh exit 8, 13, 14, 15
Bank(s) active H L DES Active power-down entry 8, 10, 12, 16
Reading H L DES Power-down entry 8, 10, 12, 16, 17
Writing H L DES Power-down entry 8, 10, 12, 16, 17
Precharging H L DES Power-down entry 8, 10, 12, 16, 17
Refreshing H L DES Precharge power-down entry 8, 12
All banks idle H L DES Precharge power-down entry 8, 10, 12, 16, 18
H L REFRESH Self refresh 16, 18, 19
Notes: 1. Current state is defined as the state of the DDR4 SDRAM immediately prior to clock
edge n.
2. CKE (n) is the logic state of CKE at clock edge n; CKE (n-1) was the state of CKE at the
previous clock edge.
3. COMMAND (n) is the command registered at clock edge n, and ACTION (n) is a result of
COMMAND (n); ODT is not included here.
4. All states and sequences not shown are illegal or reserved unless explicitly described
elsewhere in this document.
5. The state of ODT does not affect the states described in this table. The ODT function is
not available during self refresh.
6. During any CKE transition (registration of CKE H->L or CKE H->L), the CKE level must be
maintained until 1 nCK prior to tCKE (MIN) being satisfied (at which time CKE may tran-
sition again).
7. DESELECT and NOP are defined in the Truth Table – Command table.
8. For power-down entry and exit parameters, see the Power-Down Modes section.
9. CKE LOW is allowed only if tMRD and tMOD are satisfied.
10. The power-down mode does not perform any REFRESH operations.
11. X = "Don’t Care" (including floating around VREF) in self refresh and power-down. X al-
so applies to address pins.
12. The DESELECT command is the only valid command for power-down entry and exit.
13. VPP and VREFCA must be maintained during SELF REFRESH operation.
14. On self refresh exit, the DESELECT command must be issued on every clock edge occur-
ring during the tXS period. READ or ODT commands may be issued only after tXSDLL is
satisfied.
15. The DESELECT command is the only valid command for self refresh exit.
16. Self refresh cannot be entered during READ or WRITE operations. For a detailed list of
restrictions see the SELF REFRESH Operation and Power-Down Modes sections.
17. If all banks are closed at the conclusion of the READ, WRITE, or PRECHARGE command,
then precharge power-down is entered; otherwise, active power-down is entered.
NOP Command
18. Idle state is defined as all banks are closed (tRP, tDAL, and so on, satisfied), no data
bursts are in progress, CKE is HIGH, and all timings from previous operations are satis-
fied (tMRD, tMOD, tRFC, tZQinit, tZQoper, tZQCS, and so on), as well as all self refresh ex-
it and power-down exit parameters are satisfied (tXS, tXP, tXSDLL, and so on).
19. Self refresh mode can be entered only from the all banks idle state.
20. For more details about all signals, see the Truth Table – Command table; must be a legal
command as defined in the table.
NOP Command
The NO OPERATION (NOP) command was originally used to instruct the selected
DDR4 SDRAM to perform a NOP (CS_n = LOW and ACT_n, RAS_n/A16, CAS_n/A15, and
WE_n/A14 = HIGH). This prevented unwanted commands from being registered during
idle or wait states. NOP command general support has been removed and the com-
mand should not be used unless specifically allowed, which is when exiting maximum
power-saving mode or when entering gear-down mode.
DESELECT Command
The deselect function (CS_n HIGH) prevents new commands from being executed;
therefore, with this command, the device is effectively deselected. Operations already in
progress are not affected.
DLL-Off Mode
DLL-off mode is entered by setting MR1 bit A0 to 0, which will disable the DLL for sub-
sequent operations until the A0 bit is set back to 1. The MR1 A0 bit for DLL control can
be switched either during initialization or during self refresh mode. Refer to the Input
Clock Frequency Change section for more details.
The maximum clock frequency for DLL-off mode is specified by the parameter
tCKDLL_OFF.
Due to latency counter and timing restrictions, only one CL value and CWL value (in
MR0 and MR2 respectively) are supported. The DLL-off mode is only required to sup-
port setting both CL = 10 and CWL = 9.
DLL-off mode will affect the read data clock-to-data strobe relationship (tDQSCK), but
not the data strobe-to-data relationship (tDQSQ, tQH). Special attention is needed to
line up read data to the controller time domain.
Compared with DLL-on mode, where tDQSCK starts from the rising clock edge (AL +
CL) cycles after the READ command, the DLL-off mode tDQSCK starts (AL + CL - 1) cy-
cles after the READ command. Another difference is that tDQSCK may not be small
compared to tCK (it might even be larger than tCK), and the difference between tDQSCK
(MIN) and tDQSCK (MAX) is significantly larger than in DLL-on mode. The tDQSCK
(DLL-off) values are undefined and the user is responsible for training to the data-eye.
The timing relations on DLL-off mode READ operation are shown in the following dia-
gram, where CL = 10, AL = 0, and BL = 8.
DLL-Off Mode
Figure 15: DLL-Off Mode Read Timing Operation

T0 (( T1 T6 T7 T8 T9 T10 T11 T12 T13 T14
CK_c ))
((
))
CK_t ((
))
((
))
Command RD (( DES DES DES DES DES DES DES DES DES DES
))
((
))
((
))
((
Address A RD ))
((
)) RL (DLL-on) = AL + CL = 10
(( CL = 10, AL = 0
DQS_t, DQS_c ))
(DLL-on)
tDQSCK (MIN) tDQSCK (MAX)
((
DQS_c )) DIN DIN DIN DIN DIN DIN DIN DIN
(DLL-on) b b+1 b+2 b+3 b+4 b+5 b+6 b+7
RL (DLL-off) = AL + (CL - 1) = 9
CL = 10, AL = 0 tDQSCK (DLL-off) MIN
DQS_t, DQS_c ((
))
(DLL-off)
DQS_c ((
)) DIN DIN DIN DIN DIN DIN DIN DIN
(DLL-off)
b b+1 b+2 b+3 b+4 b+5 b+6 b+7
tDQSCK (DLL-off) MAX

DQS_t, DQS_c ((
(DLL-off) ))
DQS_c ((
(DLL-off) )) DIN DIN DIN DIN DIN DIN DIN DIN
b b+1 b+2 b+3 b+4 b+5 b+6 b+7
Transitioning data Don’t Care
DLL-On/Off Switching Procedures

The DLL-off mode is entered by setting MR1 bit A0 to 0; this will disable the DLL for
subsequent operations until the A0 bit is set back to 1.
DLL Switch Sequence from DLL-On to DLL-Off

To switch from DLL-on to DLL-off requires the frequency to be changed during self re-
fresh, as outlined in the following procedure:
1. Starting from the idle state (all banks pre-charged, all timings fulfilled, and, to dis-
able the DLL, the DRAM on-die termination resistors, RTT(NOM), must be in High-Z
before MRS to MR1.)
2. Set MR1 bit A0 to 1 to disable the DLL.
3. Wait tMOD.
4. Enter self refresh mode; wait until tCKSRE/tCKSRE_PAR is satisfied.
5. Change frequency, following the guidelines in the Input Clock Frequency Change
section.
6. Wait until a stable clock is available for at least tCKSRX at device inputs.
7. Starting with the SELF REFRESH EXIT command, CKE must continuously be regis-
tered HIGH until all tMOD timings from any MRS command are satisfied. In addi-
tion, if any ODT features were enabled in the mode registers when self refresh
mode was entered, the ODT signal must continuously be registered LOW until all
tMOD timings from any MRS command are satisfied. If R
TT(NOM) was disabled in
the mode registers when self refresh mode was entered, the ODT signal is "Don't
Care."
8. Wait tXS_FAST, tXS_ABORT, or tXS, and then set mode registers with appropriate
values (an update of CL, CWL, and WR may be necessary; a ZQCL command can
also be issued after tXS_FAST).
• tXS_FAST: ZQCL, ZQCS, and MRS commands. For MRS commands, only CL and
WR/RTP registers in MR0, the CWL register in MR2, and gear-down mode in
MR3 may be accessed provided the device is not in per-DRAM addressability
mode. Access to other device mode registers must satisfy tXS timing.
• tXS_ABORT: If MR4 [9] is enabled, then the device aborts any ongoing refresh
and does not increment the refresh counter. The controller can issue a valid
command after a delay of tXS_ABORT. Upon exiting from self refresh, the device
requires a minimum of one extra REFRESH command before it is put back into
self refresh mode. This requirement remains the same regardless of the MRS bit
setting for self refresh abort.
t
• XS: ACT, PRE, PREA, REF, SRE, PDE, WR, WRS4, WRS8, WRA, WRAS4, WRAS8,
RD, RDS4, RDS8, RDA, RDAS4, and RDAS8.
9. Wait tMOD to complete.
The device is ready for the next command.
Figure 16: DLL Switch Sequence from DLL-On to DLL-Off
Ta Tb0 Tb1 Tc Td Te0 Te1 Tf Tg Th

CK_c
CK_t
tCKSRE/tCKSRE_PAR Note 4 tCKSRX5
tIS
tCPDED
CKE Valid Valid Valid
tXS_FAST
Command MRS2 SRE3 DES SRX6 Valid 7 Valid 8 Valid 9
Address Valid Valid Valid

tRP tXS_ABORT
tIS tCKESR/tCKESR_PAR tXS
ODT Valid
Enter self refresh Exit self refresh
Notes: 1. Starting in the idle state. RTT in stable state.

2. Disable DLL by setting MR1 bit A0 to 0.
3. Enter SR.
4. Change frequency.
5. Clock must be stable tCKSRX.
6. Exit SR.
7. Update mode registers allowed with DLL-off settings met.
DLL-Off to DLL-On Procedure

To switch from DLL-off to DLL-on (with required frequency change) during self refresh:
1. Starting from the idle state (all banks pre-charged, all timings fulfilled, and DRAM
ODT resistors (RTT(NOM)) must be in High-Z before self refresh mode is entered.)
2. Enter self refresh mode; wait until tCKSRE/tCKSRE_PAR are satisfied.
3. Change frequency (following the guidelines in the Input Clock Frequency Change
section).
4. Wait until a stable clock is available for at least tCKSRX at device inputs.
5. Starting with the SELF REFRESH EXIT command, CKE must continuously be regis-
tered HIGH until tDLLK timing from the subsequent DLL RESET command is sat-
isfied. In addition, if any ODT features were enabled in the mode registers when
self refresh mode was entered, the ODT signal must continuously be registered
LOW or HIGH until tDLLK timing from the subsequent DLL RESET command is
satisfied. If RTT(NOM) disabled in the mode registers when self refresh mode was
entered, the ODT signal is "Don't Care."
6. Wait tXS or tXS_ABORT, depending on bit 9 in MR4, then set MR1 bit A0 to 0 to en-
able the DLL.
7. Wait tMRD, then set MR0 bit A8 to 1 to start DLL reset.
8. Wait tMRD, then set mode registers with appropriate values; an update of CL,
CWL, and WR may be necessary. After tMOD is satisfied from any proceeding MRS
command, a ZQCL command can also be issued during or after tDLLK.
9. Wait for tMOD to complete. Remember to wait tDLLK after DLL RESET before ap-
plying any command requiring a locked DLL. In addition, wait for tZQoper in case
a ZQCL command was issued.
The device is ready for the next command.
Figure 17: DLL Switch Sequence from DLL-Off to DLL-On
Ta Tb0 Tb1 Tc Td Te0 Te1 Tf Tg Th

CK_c
CK_t
Note 1 tCKSRE/tCKSRE_PAR Note 4 tCKSRX5
tIS
tCPDED
tXS_ABORT
Command MRS2 SRE3 DES SRX6 Valid 7 Valid 7 Valid 7

tRP tXS tMRD
tIS
tCKESR/tCKESR_PAR
ODT Valid
Enter self refresh Exit self refresh
Notes: 1. Starting in the idle state.
Input Clock Frequency Change
2. Enter SR.
3. Change frequency.
4. Clock must be stable tCKSRX.
5. Exit SR.
6. Set DLL to on by setting MR1 to A0 = 0.
7. Update mode registers.
8. Issue any valid command.
Input Clock Frequency Change

After the device is initialized, it requires the clock to be stable during almost all states of
normal operation. This means that after the clock frequency has been set and is in the
stable state, the clock period is not allowed to deviate except for what is allowed by the
clock jitter and spread spectrum clocking (SSC) specifications. The input clock frequen-
cy can be changed from one stable clock rate to another stable clock rate only when in
self refresh mode. Outside of self refresh mode, it is illegal to change the clock frequen-
cy.
After the device has been successfully placed in self refresh mode and tCKSRE/
tCKSRE_PAR have been satisfied, the state of the clock becomes a "Don’t Care." Follow-
ing a "Don’t Care," changing the clock frequency is permissible, provided the new clock
frequency is stable prior to tCKSRX. When entering and exiting self refresh mode for the
sole purpose of changing the clock frequency, the self refresh entry and exit specifica-
tions must still be met as outlined in SELF REFRESH Operation.
For the new clock frequency, additional MRS commands to MR0, MR2, MR3, MR4, MR5,
and MR6 may need to be issued to program appropriate CL, CWL, gear-down mode,
READ and WRITE preamble, Command Address Latency, and tCCD_L/tDLLK values.
When the clock rate is being increased (faster), the MR settings that require additional
clocks should be updated prior to the clock rate being increased. In particular, the PL
latency must be disabled when the clock rate changes, ie. while in self refresh mode. For
example, if changing the clock rate from DDR4-2133 to DDR4-2933 with CA parity
mode enabled, MR5[2:0] must first change from PL = 4 to PL = disable prior to PL = 6.
The correct procedure would be to (1) change PL = 4 to disable via MR5 [2:0], (2) enter
self refresh mode, (3) change clock rate from DDR4-2133 to DDR4-2933, (4) exit self re-
fresh mode, (5) Enable CA parity mode setting PL = 6 vis MR5 [2:0].
If the MR settings that require additional clocks are updated after the clock rate has
been increased, for example. after exiting self refresh mode, the required MR settings
must be updated prior to removing the DRAM from the IDLE state, unless the DRAM is
RESET. If the DRAM leaves the IDLE state to enter self refresh mode or ZQ Calibration,
the updating of the required MR settings may be deferred to the next time the DRAM
enters the IDLE state.
If MR6 is issued prior to self refresh entry for new the tDLLK value, DLL will relock auto-
matically at self refresh exit. However, if MR6 is issued after self refresh entry, MR0 must
be issued to reset the DLL.
The device input clock frequency can change only within the minimum and maximum
operating frequency specified for the particular speed grade. Any frequency change be-
low the minimum operating frequency would require the use of DLL-on mode to DLL-
off mode transition sequence (see DLL-On/Off Switching Procedures).
Write Leveling
Write Leveling
For better signal integrity, DDR4 memory modules use fly-by topology for the com-
mands, addresses, control signals, and clocks. Fly-by topology has benefits from the re-
duced number of stubs and their length, but it also causes flight-time skew between
clock and strobe at every DRAM on the DIMM. This makes it difficult for the controller
to maintain tDQSS, tDSS, and tDSH specifications. Therefore, the device supports a
write leveling feature to allow the controller to compensate for skew. This feature may
not be required under some system conditions, provided the host can maintain the
tDQSS, tDSS, and tDSH specifications.
The memory controller can use the write leveling feature and feedback from the device
to adjust the DQS (DQS_t, DQS_c) to CK (CK_t, CK_c) relationship. The memory con-
troller involved in the leveling must have an adjustable delay setting on DQS to align the
rising edge of DQS with that of the clock at the DRAM pin. The DRAM asynchronously
feeds back CK, sampled with the rising edge of DQS, through the DQ bus. The controller
repeatedly delays DQS until a transition from 0 to 1 is detected. The DQS delay estab-
lished though this exercise would ensure the tDQSS specification. Besides tDQSS, tDSS
and tDSH specifications also need to be fulfilled. One way to achieve this is to combine
the actual tDQSS in the application with an appropriate duty cycle and jitter on the DQS
signals. Depending on the actual tDQSS in the application, the actual values for tDQSL
and tDQSH may have to be better than the absolute limits provided in the AC Timing
Parameters section in order to satisfy tDSS and tDSH specifications. A conceptual tim-
ing of this scheme is shown below.
Figure 18: Write Leveling Concept, Example 1

T0 T1 T2 T3 T4 T5 T6 T7
CK_c
Source
CK_t
diff_DQS
Tn T0 T1 T2 T3 T4 T5 T6
CK_c
Destination
CK_t
diff_DQS
DQ 0 or 1 0 0 0
Push DQS to capture

diff_DQS the 0-1 transition
DQ 0 or 1 1 1 1
DQS driven by the controller during leveling mode must be terminated by the DRAM
based on the ranks populated. Similarly, the DQ bus driven by the DRAM must also be
terminated at the controller.
All data bits carry the leveling feedback to the controller across the DRAM configura-
tions: x4, x8, and x16. On a x16 device, both byte lanes should be leveled independently.
Therefore, a separate feedback mechanism should be available for each byte lane. The
upper data bits should provide the feedback of the upper diff_DQS(diff_UDQS)-to-
clock relationship; the lower data bits would indicate the lower diff_DQS(diff_LDQS)-
to-clock relationship.
Write Leveling
The figure below is another representative way to view the write leveling procedure. Al-
though it shows the clock varying to a static strobe, this is for illustrative purpose only;
the clock does not actually change phase, the strobe is what actually varies. By issuing
multiple WL bursts, the DQS strobe can be varied to capture with fair accuracy the time
at which the clock edge arrives at the DRAM clock input buffer.
Figure 19: Write Leveling Concept, Example 2

tWLS
CK_c
1 1 1 111 1 1 11 1 1 1 1 1 1 1 1
CK_t
tWLH
CK_c
0 000 000 000 000
CK_t
tWLS tWLH
CK_c
CK_t
0 000 000 X XX X X X 11 1 1 1 1 1 1 1 1
DQS_t/
DQS_c
tWLO
DQ (CK 0 to 1)
DQ (CK 1 to 0)
DRAM Setting for Write Leveling and DRAM TERMINATION Function in that Mode
The DRAM enters into write leveling mode if A7 in MR1 is HIGH. When leveling is fin-
ished, the DRAM exits write leveling mode if A7 in MR1 is LOW (see the MR Leveling
Procedures table). Note that in write leveling mode, only DQS terminations are activa-
ted and deactivated via the ODT pin, unlike normal operation (see DRAM DRAM TER-
MINATION Function in Leveling Mode table).
Table 25: MR Settings for Leveling Procedures
Function MR1 Enable Disable

Write leveling enable A7 1 0
Output buffer mode (Q off) A12 0 1
Table 26: DRAM TERMINATION Function in Leveling Mode
ODT Pin at DRAM DQS_t/DQS_c Termination DQ Termination

RTT(NOM) with ODT HIGH On Off
Write Leveling
Table 26: DRAM TERMINATION Function in Leveling Mode (Continued)
ODT Pin at DRAM DQS_t/DQS_c Termination DQ Termination

RTT(Park) with ODT LOW On Off
Notes: 1. In write leveling mode, with the mode's output buffer either disabled (MR1[bit7] = 1
and MR1[bit12] = 1) or with its output buffer enabled (MR1[bit7] = 1 and MR1[bit12] =
0), all RTT(NOM) and RTT(Park) settings are supported.
2. RTT(WR) is not allowed in write leveling mode and must be set to disable prior to enter-
ing write leveling mode.
Procedure Description
The memory controller initiates the leveling mode of all DRAM by setting bit 7 of MR1
to 1. When entering write leveling mode, the DQ pins are in undefined driving mode.
During write leveling mode, only the DESELECT command is supported, other than
MRS commands to change the Qoff bit (MR1[A12]) and to exit write leveling (MR1[A7]).
Upon exiting write leveling mode, the MRS command performing the exit (MR1[A7] = 0)
may also change the other MR1 bits. Because the controller levels one rank at a time,
the output of other ranks must be disabled by setting MR1 bit A12 to 1. The controller
may assert ODT after tMOD, at which time the DRAM is ready to accept the ODT signal,
unless DODTLon or DODTLoff have been altered (the ODT internal pipe delay is in-
creased when increasing WRITE latency [WL] or READ latency [RL] by the previous MR
command), then ODT assertion should be delayed by DODTLon after tMOD is satisfied,
which means the delay is now tMOD + DODTLon.
The controller may drive DQS_t LOW and DQS_c HIGH after a delay of tWLDQSEN, at
which time the DRAM has applied ODT to these signals. After tDQSL and tWLMRD, the
controller provides a single DQS_t, DQS_c edge, which is used by the DRAM to sample
CK driven from the controller. tWLMRD (MAX) timing is controller dependent.
The DRAM samples CK status with the rising edge of DQS and provides feedback on all
the DQ bits asynchronously after tWLO timing. There is a DQ output uncertainty of
tWLOE defined to allow mismatch on DQ bits. The tWLOE period is defined from the
transition of the earliest DQ bit to the corresponding transition of the latest DQ bit.
There are no read strobes (DQS_t, DQS_c) needed for these DQs. The controller sam-
ples incoming DQ and either increments or decrements DQS delay setting and launch-
es the next DQS pulse after some time, which is controller dependent. After a 0-to-1
transition is detected, the controller locks the DQS delay setting, and write leveling is
achieved for the device. The following figure shows the timing diagram and parameters
for the overall write leveling procedure.
Write Leveling
Figure 20: Write Leveling Sequence (DQS Capturing CK LOW at T1 and CK HIGH at T2)
T1 T2
tWLH tWLH
tWLS tWLS
CK_c5
CK_t
Command MRS2 DES3 DES DES DES DES DES NOP DES DES DES DES DES
tMOD
ODT
tWLDQSEN tDQSL6 tDQSH6 tDQSL6 tDQSH6
diff_DQS4
tWLMRD
tWLO tWLO
Late Prime DQ1

tWLOE
Early Prime DQ1

tWLO tWLO
tWLOE
Undefined Driving Mode Time Break Don’t Care
Notes: 1. The device drives leveling feedback on all DQs.

2. MRS: Load MR1 to enter write leveling mode.
3. diff_DQS is the differential data strobe. Timing reference points are the zero crossings.
DQS_t is shown with a solid line; DQS_c is shown with a dotted line.
4. CK_t is shown with a solid dark line; CK_c is shown with a dotted line.
5. DQS needs to fulfill minimum pulse width requirements, tDQSH (MIN) and tDQSL (MIN),
as defined for regular WRITEs; the maximum pulse width is system dependent.
6. tWLDQSEN must be satisfied following equation when using ODT:
• DLL = Enable, then tWLDQSEN > tMOD (MIN) + DODTLon + tADC
• DLL = Disable, then tWLDQSEN > tMOD (MIN) + tAONAS
Write Leveling Mode Exit

Write leveling mode should be exited as follows:
1. After the last rising strobe edge (see ~T0), stop driving the strobe signals (see
~Tc0). Note that from this point on, DQ pins are in undefined driving mode and
will remain undefined, until tMOD after the respective MR command (Te1).
2. Drive ODT pin LOW (tIS must be satisfied) and continue registering LOW (see
Tb0).
3. After RTT is switched off, disable write leveling mode via the MRS command (see
Tc2).
4. After tMOD is satisfied (Te1), any valid command can be registered. (MR com-
mands can be issued after tMRD [Td1]).
Write Leveling
Figure 21: Write Leveling Exit

T0 T1 T2 Ta0 Tb0 Tc0 Tc1 Tc2 Td0 Td1 Te0 Te1
CK_c
CK_t
Command DES DES DES DES DES DES DES DES DES Valid DES Valid
tMRD
Address MR1 Valid Valid

tIS tMOD
ODT
ODTL (OFF) tADC (MIN)

RTT(DQS_t)
RTT(NON) RTT(Park)
RTT(DQS_c)
tADC (MAX)
DQS_t,
DQS_c
RTT(DQ)
tWLO
DQ1 result = 1
Undefined Driving Mode Transitioning Time Break Don’t Care
Notes: 1. The DQ result = 1 between Ta0 and Tc0 is a result of the DQS signals capturing CK_t
HIGH just after the T0 state.
2. See previous figure for specific tWLO timing.

DDR4 supports the command address latency (CAL) function as a power savings fea-
ture. This feature can be enabled or disabled via the MRS setting. CAL timing is defined
as the delay in clock cycles (tCAL) between a CS_n registered LOW and its correspond-
ing registered command and address. The value of CAL in clocks must be programmed
into the mode register (see MR1 Register Definition table) and is based on the tCAL(ns)/
tCK(ns) rounding algorithms found in the Converting Time-Based Specifications to
Clock-Based Requirements section.
Figure 22: CAL Timing Definition
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
CLK
CS_n
CMD/ADDR
tCAL
CAL gives the DRAM time to enable the command and address receivers before a com-
mand is issued. After the command and the address are latched, the receivers can be
disabled if CS_n returns to HIGH. For consecutive commands, the DRAM will keep the
command and address input receivers enabled for the duration of the command se-
quence.
Figure 23: CAL Timing Example (Consecutive CS_n = LOW)
1 2 3 4 5 6 7 8 9 10 11 12
CLK
CS_n
CMD/ADDR
When the CAL mode is enabled, additional time is required for the MRS command to
complete. The earliest the next valid command can be issued is tMOD_CAL, which
should be equal to tMOD + tCAL. The two following figures are examples.
Figure 24: CAL Enable Timing – tMOD_CAL

T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Tb0 Tb1 Tb2 Tb3
CK_c
CK_t
Command Valid MRS DES DES DES DES DES DES DES Valid Valid
CS_n
tCAL
tMOD_CAL
Note: 1. CAL mode is enabled at T1.
Figure 25: tMOD_CAL, MRS to Valid Command Timing with CAL Enabled
T0 T1 Ta0 Ta1 Ta2 Tb0 Tb1 Tb2 Tc0 Tc1 Tc2
CK_c
CK_t
Command Valid DES DES MRS DES DES DES DES DES Valid Valid
tCAL tCAL
CS_n
tMOD_CAL
Note: 1. MRS at Ta1 may or may not modify CAL, tMOD_CAL is computed based on new tCAL set-
ting if modified.
When the CAL mode is enabled or being enabled, the earliest the next MRS command
can be issued is tMRD_CAL is equal to tMOD + tCAL. The two following figures are ex-
amples.
Figure 26: CAL Enabling MRS to Next MRS Command, tMRD_CAL

T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Tb0 Tb1 Tb2 Tb3
CK_c
CK_t
Command Valid MRS DES DES DES DES DES DES DES MRS DES
tCAL
CS_n
tMRD_CAL
Settings Old settings Updating settings Updating settings
Note: 1. Command address latency mode is enabled at T1.
Figure 27: tMRD_CAL, Mode Register Cycle Time With CAL Enabled
7 7 7D 7D 7D 7E 7E 7E 7F 7F 7F
CK_c
CK_t
Command 9DOLG '(6 '(6 056 '(6 '(6 '(6 '(6 '(6 056 '(6
W &$/ W &$/
Address 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG
CS_n
W 05'B&$/
Settings 2OGVHWWLQJV 8SGDWLQJVHWWLQJV 1HZVHWWLQJV
7LPH%UHDN 'RQ¶W&DUH
Note: 1. MRS at Ta1 may or may not modify CAL, tMRD_CAL is computed based on new tCAL set-
ting if modified.
CAL Examples: Consecutive READ BL8 with two different CALs and 1tCK preamble in
different bank group shown in the following figures.
CCMTD-1725822587-10418
Figure 28: Consecutive READ BL8, CAL3, 1tCK Preamble, Different Bank Group
T0 T1 T2 T3 T4 T5 T6 T7 T13 T14 T15 T16 T17 T18 T19 T20 T21 T22
CK_c
CK_t
CS_n
t t
CAL = 3 CAL = 3
Command DES DES READ DES DES READ DES DES DES DES DES DES DES DES DES DES
t
CCD_S = 4
Bank Group BG a BG b
Address
Bank, Bank,
Address Col n Col b
tRPST
tRPRE (1nCK)
DQS_t, DQS_c
DQ DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT
RL = 11 n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+7 b+2 b+3 b+4 b+5 b+6 b+7
RL = 11
Transitioning Data Don’t Care
Notes: 1. BL = 8, AL = 0, CL = 11, CAL = 3, Preamble = 1tCK.

2. DOUT n = data-out from column n; DOUT b = data-out from column b.
3. DES commands are shown for ease of illustration, other commands may be valid at these times.
4. BL8 setting activated by either MR0[A1:0 = 00] or MR0[A1:0 = 01] and A12 = 1 during READ command at T3 and
T7.
5. CA parity = Disable, CS to CA latency = Enable, Read DBI = Disable.
6. Enabling CAL mode does not impact ODT control timings. ODT control timings should be maintained with the
84
same timing relationship relative to the command/address bus as when CAL is disabled.

Figure 29: Consecutive READ BL8, CAL4, 1tCK Preamble, Different Bank Group
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
&.BF
&.BW
&6BQ
W
&$/ W
&$/

&RPPDQG '(6 '(6 '(6 5($' '(6 '(6 '(6 5($' '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6
W
&&'B6
%DQN*URXS %*D %*E
$GGUHVV
%DQN %DQN
$GGUHVV &ROQ &ROE
W 5367
W 535(Q&.
'46BW'46BF
'4 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287 '287
5/ Q Q Q Q Q Q Q Q E E E E E E E E
5/
7UDQVLWLRQLQJ'DWD 'RQ¶W&DUH
Notes: 1. BL = 8, AL = 0, CL = 11, CAL = 4, Preamble = 1tCK.

2. DOUT n = data-out from column n; DOUT b = data-out from column b.
3. DES commands are shown for ease of illustration, other commands may be valid at these times.
4. BL8 setting activated by either MR0[A1:0 = 00] or MR0[A1:0 = 01] and A12 = 1 during READ command at T4 and
T8.
6. Enabling CAL mode does not impact ODT control timings. ODT control timings should be maintained with the
same timing relationship relative to the command/address bus as when CAL is disabled.
5. CA parity = Disable, CS to CA latency = Enable, Read DBI = Disable.
Low-Power Auto Self Refresh Mode

An auto self refresh mode is provided for application ease. Auto self refresh mode is en-
abled by setting MR2[6] = 1 and MR2[7] = 1. The device will manage self refresh entry
over the supported temperature range of the DRAM. In this mode, the device will
change its self refresh rate as the DRAM operating temperature changes, going lower at
low temperatures and higher at high temperatures.
Manual Self Refresh Mode

If auto self refresh mode is not enabled, the low-power auto self refresh mode register
must be manually programmed to one of the three self refresh operating modes. This
mode provides the flexibility to select a fixed self refresh operating mode at the entry of
the self refresh, according to the system memory temperature conditions. The user is
responsible for maintaining the required memory temperature condition for the mode
selected during the SELF REFRESH operation. The user may change the selected mode
after exiting self refresh and before entering the next self refresh. If the temperature
condition is exceeded for the mode selected, there is a risk to data retention resulting in
loss of data.
Table 27: Auto Self Refresh Mode
Low-Power Operating Temperature

Auto Self Refresh Range for Self Refresh Mode
MR2[7] MR2[6] Mode SELF REFRESH Operation (DRAM TCASE)
0 0 Normal Variable or fixed normal self refresh rate –40°C to 85°C
maintains data retention at the normal oper-
ating temperature. User is required to ensure
that 85°C DRAM TCASE (MAX) is not exceeded
to avoid any risk of data loss.
1 0 Extended Variable or fixed high self refresh rate opti- –40°C to 125°C
temperature mizes data retention to support the exten-
ded temperature range.
0 1 Reduced Variable or fixed self refresh rate or any oth- –40°C to 45°C
temperature er DRAM power consumption reduction con-
trol for the reduced temperature range. User
is required to ensure 45°C DRAM TCASE
(MAX) is not exceeded to avoid any risk of
data loss.
1 1 Auto self refresh Auto self refresh mode enabled. Self refresh All of the above
power consumption and data retention are
optimized for any given operating tempera-
ture condition.
Figure 30: Auto Self Refresh Ranges

IDD6
2x refresh rate
1x refresh rate
Extended
temperature
range
1/2x refresh rate
Reduced Normal
temperature temperature
range range
Tc
-40°C 45°C 85°C 105°C
The MULTIPURPOSE REGISTER (MPR) function, MPR access mode, is used to write/
read specialized data to/from the DRAM. The MPR consists of four logical pages, MPR
Page 0 through MPR Page 3, with each page having four 8-bit registers, MPR0 through
MPR3. Page 0 can be read by any of three readout modes (serial, parallel, or staggered)
while Pages 1, 2, and 3 can be read by only the serial readout mode. Page 3 is for DRAM
vendor use only. MPR mode enable and page selection is done with MRS commands.
Data bus inversion (DBI) is not allowed during MPR READ operation.
Once the MPR access mode is enabled (MR3[2] = 1), only the following commands are
allowed: MRS, RD, RDA WR, WRA, DES, REF, and RESET; RDA/WRA have the same func-
tionality as RD/WR which means the auto precharge part of RDA/WRA is ignored. Pow-
er-down mode and SELF REFRESH command are not allowed during MPR enable
mode. No other command can be issued within tRFC after a REF command has been
issued; 1x refresh (only) is to be used during MPR access mode. While in MPR access
mode, MPR read or write sequences must be completed prior to a REFRESH command.
Figure 31: MPR Block Diagram
Memory core
(all banks precharged)
Four multipurpose registers (pages),

MR3 [2] = 1 each with four 8-bit registers:
Data patterns (RD/WR)

Error log (RD)
Mode registers (RD)
flow
data DRAM manufacture only (RD)
PR
M
DQ,s DM_n/DBI_n, DQS_t, DQS_c
Table 28: MR3 Setting for the MPR Access Mode
Address Operation Mode Description

A[12:11] MPR data read format 00 = Serial ........... 01 = Parallel
10 = Staggered .... 11 = Reserved
A2 MPR access 0 = Standard operation (MPR not enabled)
1 = MPR data flow enabled
A[1:0] MPR page selection 00 = Page 0 .... 01 = Page 1
10 = Page 2 .... 11 = Page 3
Table 29: DRAM Address to MPR UI Translation
MPR Location [7] [6] [5] [4] [3] [2] [1] [0]
DRAM address – Ax A7 A6 A5 A4 A3 A2 A1 A0
MPR UI – UIx UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7
Table 30: MPR Page and MPRx Definitions
Address MPR Location [7] [6] [5] [4] [3] [2] [1] [0] Note
MPR Page 0 – Read or Write (Data Patterns)
BA[1:0] 00 = MPR0 0 1 0 1 0 1 0 1 Read/
01 = MPR1 0 0 1 1 0 0 1 1 Write
(default
10 = MPR2 0 0 0 0 1 1 1 1
value lis-
11 = MPR3 0 0 0 0 0 0 0 0 ted)
MPR Page 1 – Read-only (Error Log)
BA[1:0] 00 = MPR0 A7 A6 A5 A4 A3 A2 A1 A0 Read-on-
01 = MPR1 CAS_n/A WE_n/A1 A13 A12 A11 A10 A9 A8 ly
15 4
10 = MPR2 PAR ACT_n BG1 BG0 BA1 BA0 A17 RAS_n/A
16
11 = MPR3 CRC er- CA pari- CA parity latency: [5] = C2 C1 C0
ror sta- ty error MR5[2], [4] = MR5[1], [3] =
tus status MR5[0]
MPR Page 2 – Read-only (MRS Readout)
BA[1:0] 00 = MPR0 hPPR sPPR RTT(WR) Temperature sen- CRC write RTT(WR) MR2[10:9] Read-on-
support support MR2[11] sor status2 enable ly
MR2[12]
01 = MPR1 VREFDQ VREFDQ training value: [6:1] = MR6[5:0] Gear-
traing- down
ing enable
range MR3[3]
MR6[6]
10 = MPR2 CAS latency: [7:3] = MR0[6:4,2,12] CAS write latency [2:0] =
MR2[5:3]
11 = MPR3 RTT(NOM): [7:5] = MR1[10:8] RTT(Park): [4:2] = MR5[8:6] RON: [1:0] =
MR2[2:1]
MPR Page 3 – Read-only (Restricted, except for MPR3 [3:0])
BA[1:0] 00 = MPR0 DC DC DC DC DC DC DC DC Read-on-
01 = MPR1 DC DC DC DC DC DC DC DC ly
10 = MPR2 DC DC DC DC DC DC DC DC
11 = MPR3 DC DC DC DC MAC MAC MAC MAC
Notes: 1. DC = "Don't Care"

2. MPR[4:3] 00 = Sub 1X refresh; MPR[4:3] 01 = 1X refresh; MPR[4:3] 10 = 2X refresh;
MPR[4:3] 11 = Reserved
MPR Reads
MPR reads are supported using BL8 and BC4 modes. Burst length on-the-fly is not sup-
ported for MPR reads. Data bus inversion (DBI) is not allowed during MPR READ opera-
tion; the device will ignore the Read DBI enable setting in MR5 [12] when in MPR mode.
READ commands for BC4 are supported with a starting column address of A[2:0] = 000
or 100. After power-up, the content of MPR Page 0 has the default values, which are de-
fined in Table 30. MPR page 0 can be rewritten via an MPR WRITE command. The de-
vice maintains the default values unless it is rewritten by the DRAM controller. If the
DRAM controller does overwrite the default values (Page 0 only), the device will main-
tain the new values unless re-initialized or there is power loss.
Timing in MPR mode:
• Reads (back-to-back) from Page 0 may use tCCD_S or tCCD_L timing between READ
commands
• Reads (back-to-back) from Pages 1, 2, or 3 may not use tCCD_S timing between READ
commands; tCCD_L must be used for timing between READ commands
The following steps are required to use the MPR to read out the contents of a mode reg-
ister (MPR Page x, MPRy).
1. The DLL must be locked if enabled.
2. Precharge all; wait until tRP is satisfied.
3. MRS command to MR3[2] = 1 (Enable MPR data flow), MR3[12:11] = MPR read for-
mat, and MR3[1:0] MPR page.
a. MR3[12:11] MPR read format:
1. 00 = Serial read format
2. 01 = Parallel read format
3. 10 = staggered read format
4. 11 = RFU
b. MR3[1:0] MPR page:
1. 00 = MPR Page 0
2. 01 = MPR Page 1
3. 10 = MPR Page 2
4. 11 = MPR Page 3
4. tMRD and tMOD must be satisfied.
5. Redirect all subsequent READ commands to specific MPRx location.
6. Issue RD or RDA command.
a. BA1 and BA0 indicate MPRx location:
1. 00 = MPR0
2. 01 = MPR1
3. 10 = MPR2
4. 11 = MPR3
b. A12/BC = 0 or 1; BL8 or BC4 fixed-only, BC4 OTF not supported.
1. If BL = 8 and MR0 A[1:0] = 01, A12/BC must be set to 1 during MPR
READ commands.
c. A2 = burst-type dependant:
1. BL8: A2 = 0 with burst order fixed at 0, 1, 2, 3, 4, 5, 6, 7
2. BL8: A2 = 1 not allowed
3. BC4: A2 = 0 with burst order fixed at 0, 1, 2, 3, T, T, T, T
4. BC4: A2 = 1 with burst order fixed at 4, 5, 6, 7, T, T, T, T
d. A[1:0] = 00, data burst is fixed nibble start at 00.
e. Remaining address inputs, including A10, and BG1 and BG0 are "Don’t
Care."
7. After RL = AL + CL, DRAM bursts data from MPRx location; MPR readout format
determined by MR3[A12,11,1,0].
8. Steps 5 through 7 may be repeated to read additional MPRx locations.
9. After the last MPRx READ burst, tMPRR must be satisfied prior to exiting.
10. Issue MRS command to exit MPR mode; MR3[2] = 0.
11. After the tMOD sequence is completed, the DRAM is ready for normal operation
from the core (such as ACT).
MPR Readout Format

The MPR read data format can be set to three different settings: serial, parallel, and
staggered.
MPR Readout Serial Format

The serial format is required when enabling the MPR function to read out the contents
of an MRx, temperature sensor status, and the command address parity error frame.
However, data bus calibration locations (four 8-bit registers) can be programmed to
read out any of the three formats. The DRAM is required to drive associated strobes
with the read data similar to normal operation (such as using MRS preamble settings).
Serial format implies that the same pattern is returned on all DQ lanes, as shown the
table below, which uses values programmed into the MPR via [7:0] as 0111 1111.
Table 31: MPR Readout Serial Format
Serial UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7

x4 Device
DQ0 0 1 1 1 1 1 1 1
DQ1 0 1 1 1 1 1 1 1
DQ2 0 1 1 1 1 1 1 1
DQ3 0 1 1 1 1 1 1 1
x8 Device
DQ0 0 1 1 1 1 1 1 1
DQ1 0 1 1 1 1 1 1 1
DQ2 0 1 1 1 1 1 1 1
DQ3 0 1 1 1 1 1 1 1
DQ4 0 1 1 1 1 1 1 1
DQ5 0 1 1 1 1 1 1 1
DQ6 0 1 1 1 1 1 1 1
DQ7 0 1 1 1 1 1 1 1
x16 Device
DQ0 0 1 1 1 1 1 1 1
DQ1 0 1 1 1 1 1 1 1
DQ2 0 1 1 1 1 1 1 1
DQ3 0 1 1 1 1 1 1 1
DQ4 0 1 1 1 1 1 1 1
DQ5 0 1 1 1 1 1 1 1
DQ6 0 1 1 1 1 1 1 1
DQ7 0 1 1 1 1 1 1 1
DQ8 0 1 1 1 1 1 1 1
Table 31: MPR Readout Serial Format (Continued)
Serial UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7

DQ9 0 1 1 1 1 1 1 1
DQ10 0 1 1 1 1 1 1 1
DQ11 0 1 1 1 1 1 1 1
DQ12 0 1 1 1 1 1 1 1
DQ13 0 1 1 1 1 1 1 1
DQ14 0 1 1 1 1 1 1 1
DQ15 0 1 1 1 1 1 1 1
MPR Readout Parallel Format

Parallel format implies that the MPR data is returned in the first data UI and then repea-
ted in the remaining UIs of the burst, as shown in the table below. Data pattern location
0 is the only location used for the parallel format. RD/RDA from data pattern locations
1, 2, and 3 are not allowed with parallel data return mode. In this example, the pattern
programmed in the data pattern location 0 is 0111 1111. The x4 configuration only out-
puts the first four bits (0111 in this example). For the x16 configuration, the same pat-
tern is repeated on both the upper and lower bytes.
Table 32: MPR Readout – Parallel Format
Parallel UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7

x4 Device
DQ0 0 0 0 0 0 0 0 0
DQ1 1 1 1 1 1 1 1 1
DQ2 1 1 1 1 1 1 1 1
DQ3 1 1 1 1 1 1 1 1
x8 Device
DQ0 0 0 0 0 0 0 0 0
DQ1 1 1 1 1 1 1 1 1
DQ2 1 1 1 1 1 1 1 1
DQ3 1 1 1 1 1 1 1 1
DQ4 1 1 1 1 1 1 1 1
DQ5 1 1 1 1 1 1 1 1
DQ6 1 1 1 1 1 1 1 1
DQ7 1 1 1 1 1 1 1 1
x16 Device
DQ0 0 0 0 0 0 0 0 0
DQ1 1 1 1 1 1 1 1 1
DQ2 1 1 1 1 1 1 1 1
DQ3 1 1 1 1 1 1 1 1
DQ4 1 1 1 1 1 1 1 1
Table 32: MPR Readout – Parallel Format (Continued)
Parallel UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7

DQ5 1 1 1 1 1 1 1 1
DQ6 1 1 1 1 1 1 1 1
DQ7 1 1 1 1 1 1 1 1
DQ8 0 0 0 0 0 0 0 0
DQ9 1 1 1 1 1 1 1 1
DQ10 1 1 1 1 1 1 1 1
DQ11 1 1 1 1 1 1 1 1
DQ12 1 1 1 1 1 1 1 1
DQ13 1 1 1 1 1 1 1 1
DQ14 1 1 1 1 1 1 1 1
DQ15 1 1 1 1 1 1 1 1
MPR Readout Staggered Format

Staggered format of data return is defined as the staggering of the MPR data across the
lanes. In this mode, an RD/RDA command is issued to a specific data pattern location
and then the data is returned on the DQ from each of the different data pattern loca-
tions. For the x4 configuration, an RD/RDA to data pattern location 0 will result in data
from location 0 being driven on DQ0, data from location 1 being driven on DQ1, data
from location 2 being driven on DQ2, and so on, as shown below. Similarly, an RD/RDA
command to data pattern location 1 will result in data from location 1 being driven on
DQ0, data from location 2 being driven on DQ1, data from location 3 being driven on
DQ2, and so on. Examples of different starting locations are also shown.
Table 33: MPR Readout Staggered Format, x4
x4 READ MPR0 Command x4 READ MPR1 Command x4 READ MPR2 Command x4 READ MPR3 Command
Stagger UI[7:0] Stagger UI[7:0] Stagger UI[7:0] Stagger UI[7:0]
DQ0 MPR0 DQ0 MPR1 DQ0 MPR2 DQ0 MPR3
It is expected that the DRAM can respond to back-to-back RD/RDA commands to the
MPR for all DDR4 frequencies so that a sequence (such as the one that follows) can be
created on the data bus with no bubbles or clocks between read data. In this case, the
system memory controller issues a sequence of RD(MPR0), RD(MPR1), RD(MPR2),
RD(MPR3), RD(MPR0), RD(MPR1), RD(MPR2), and RD(MPR3).
Table 34: MPR Readout Staggered Format, x4 – Consecutive READs
Stagger UI[7:0] UI[15:8] UI[23:16] UI[31:24] UI[39:32] UI[47:40] UI[55:48] UI[63:56]

DQ0 MPR0 MPR1 MPR2 MPR3 MPR0 MPR1 MPR2 MPR3
Table 34: MPR Readout Staggered Format, x4 – Consecutive READs (Continued)
Stagger UI[7:0] UI[15:8] UI[23:16] UI[31:24] UI[39:32] UI[47:40] UI[55:48] UI[63:56]

For the x8 configuration, the same pattern is repeated on the lower nibble as on the up-
per nibble. READs to other MPR data pattern locations follow the same format as the x4
case. A read example to MPR0 for x8 and x16 configurations is shown below.
Table 35: MPR Readout Staggered Format, x8 and x16
x8 READ MPR0 Command x16 READ MPR0 Command x16 READ MPR0 Command
Stagger UI[7:0] Stagger UI[7:0] Stagger UI[7:0]
DQ0 MPR0 DQ0 MPR0 DQ8 MPR0
MPR READ Waveforms

The following waveforms show MPR read accesses.
Figure 32: MPR READ Timing

T0 Ta0 Ta1 Tb0 Tc0 Tc1 Tc2 Tc3 Td0 Td1 Te0 Tf0 Tf1
CK_c
CK_t
MPE Enable MPE Disable
Command PREA MRS1 DES READ DES DES DES DES DES DES MRS3 Valid 4 DES
tRP tMOD tMPRR tMOD
Address Valid Valid Valid Add2 Valid Valid Valid Valid Valid Valid Valid Valid Valid
CKE
PL5 + AL + CL
DQS_t,
DQS_c
DQ UI0 UI1 UI2 UI5 UI6 UI7
Notes: 1. tCCD_S = 4tCK, Read Preamble = 1tCK.

2. Address setting:
A[1:0] = 00b (data burst order is fixed starting at nibble, always 00b here)
A2 = 0b (for BL = 8, burst order is fixed at 0, 1, 2, 3, 4, 5, 6, 7)

BA1 and BA0 indicate the MPR location
A10 and other address pins are "Don’t Care," including BG1 and BG0. A12 is "Don’t
Care" when MR0 A[1:0] = 00 or 10 and must be 1b when MR0 A[1:0] = 01
3. Multipurpose registers read/write disable (MR3 A2 = 0).
4. Continue with regular DRAM command.
5. Parity latency (PL) is added to data output delay when CA parity latency mode is ena-
bled.
Figure 33: MPR Back-to-Back READ Timing

T0 T1 T2 T3 T4 T5 T6 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Ta9 Ta10
CK_c
CK_t
Command DES READ DES DES DES READ DES DES DES DES DES DES DES DES DES DES DES DES
tCCD_S1
Address Valid Add2 Valid Add2 Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid
CKE
PL3 + AL + CL
DQS_t,
DQS_c
DQ UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7 UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7
DQS_t,
DQS_c
DQ UI0 UI1 UI2 UI3 UI0 UI1 UI2 UI3
Notes: 1. tCCD_S = 4tCK, Read Preamble = 1tCK.

2. Address setting:
A2 = 0b (for BL = 8, burst order is fixed at 0, 1, 2, 3, 4, 5, 6, 7; for BC = 4, burst order is
fixed at 0, 1, 2, 3, T, T, T, T)
Care" when MR0 A[1:0] = 00 or 10 and must be 1b when MR0 A[1:0] = 01
bled.
Figure 34: MPR READ-to-WRITE Timing

T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Tb0 Tb1 Tb2
CK_c
CK_t
Command READ DES DES DES DES DES DES DES DES DES WRITE DES DES
tMPRR
Address Add1 Valid Valid Valid Valid Valid Valid Valid Valid Valid Add2 Valid Valid
CKE
PL3 + AL + CL
DQS_t,
DQS_c
Notes: 1. Address setting:

Care" when MR0 A[1:0] = 00 and must be 1b when MR0 A[1:0] = 01
2. Address setting:
A[7:0] = data for MPR
A10 and other address pins are "Don’t Care"
bled.
MPR Writes
MPR access mode allows 8-bit writes to the MPR Page 0 using the address bus A[7:0].
Data bus inversion (DBI) is not allowed during MPR WRITE operation. The DRAM will
maintain the new written values unless re-initialized or there is power loss.
The following steps are required to use the MPR to write to mode register MPR Page 0.
1. The DLL must be locked if enabled.
2. Precharge all; wait until tRP is satisfied.
3. MRS command to MR3[2] = 1 (enable MPR data flow) and MR3[1:0] = 00 (MPR
Page 0); writes to 01, 10, and 11 are not allowed.
4. tMRD and tMOD must be satisfied.
5. Redirect all subsequent WRITE commands to specific MPRx location.
6. Issue WR or WRA command:
a. BA1 and BA0 indicate MPRx location
1. 00 = MPR0
2. 01 = MPR1
3. 10 = MPR2
4. 11 = MPR3
b. A[7:0] = data for MPR Page 0, mapped A[7:0] to UI[7:0].
c. Remaining address inputs, including A10, and BG1 and BG0 are "Don’t
Care."
7. tWR_MPR must be satisfied to complete MPR WRITE.
8. Steps 5 through 7 may be repeated to write additional MPRx locations.

9. After the last MPRx WRITE, tMPRR must be satisfied prior to exiting.
10. Issue MRS command to exit MPR mode; MR3[2] = 0.
11. When the tMOD sequence is completed, the DRAM is ready for normal operation
from the core (such as ACT).
MPR WRITE Waveforms

The following waveforms show MPR write accesses.
Figure 35: MPR WRITE and WRITE-to-READ Timing

T0 Ta0 Ta1 Tb0 Tc0 Tc1 Tc2 Td0 Td1 Td2 Td3 Td4 Td5
CK_c
CK_t
MPR Enable
Command PREA MRS1 DES WRITE DES DES READ DES DES DES DES DES DES
tRP tMOD tWR_MPR
Address Valid Valid Valid Add2 Valid Valid Add Valid Valid Valid Add2 Valid Valid
CKE
PL3 + AL + CL
DQS_t,
DQS_c
Notes: 1. Multipurpose registers read/write enable (MR3 A2 = 1).

2. Address setting:
bled.
Figure 36: MPR Back-to-Back WRITE Timing

T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Ta9 Ta10
CK_c
CK_t
Command WRITE DES DES DES WRITE DES DES DES DES DES DES DES DES
tWR_MPR
Address Add1 Valid Valid Add1 Valid Valid Add Valid Valid Valid Valid Valid Valid
CKE
DQS_t,
DQS_c
DQ
Note: 1. Address setting:

MPR REFRESH Waveforms

The following waveforms show MPR accesses interaction with refreshes.
Figure 37: REFRESH Timing

T0 Ta0 Ta1 Tb0 Tb1 Tb2 Tb3 Tb4 Tc0 Tc1 Tc2 Tc3 Tc4
CK_c
CK_t
MPR Enable
Command PREA MRS1 DES REF2 DES DES DES DES DES DES Valid Valid Valid
tRP tMOD tRFC
Address Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid
Notes: 1. Multipurpose registers read/write enable (MR3 A2 = 1). Redirect all subsequent read and
writes to MPR locations.
2. 1x refresh is only allowed when MPR mode is enabled.
Figure 38: READ-to-REFRESH Timing

T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Ta9
CK_c
CK_t
Command READ DES DES DES DES DES DES DES DES DES REF2 DES DES
Address Add1 Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid
CKE
PL + AL + CL (4 + 1) Clocks tRFC
BL = 8
DQS_t, DQS_c
BC = 4
DQS_t, DQS_c
DQ UI0 UI1 UI2 UI3

Care" when MR0 A[1:0] = 00 or 10, and must be 1b when MR0 A[1:0] = 01
Figure 39: WRITE-to-REFRESH Timing

T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Ta9 Ta10
CK_c
CK_t
Command WRITE DES DES DES REF2 DES DES DES DES DES DES DES DES
tWR_MPR tRFC
Address Add1 Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid
CKE
DQS_t,
DQS_c
DQ


Gear-Down Mode
Gear-Down Mode
The DDR4 SDRAM defaults in 1/2 rate (1N) clock mode and uses a low-frequency MRS
command (the MRS command has relaxed setup and hold) followed by a sync pulse
(first CS pulse after MRS setting) to align the proper clock edge for operating the control
lines CS_n, CKE, and ODT when in 1/4 rate (2N) mode. Gear-down mode is only sup-
ported at DDR4-2666 and faster. For operation in 1/2 rate mode, neither an MRS com-
mand or a sync pulse is required. Gear-down mode may only be entered during initiali-
zation or self refresh exit and may only be exited during self refresh exit. CAL mode and
CA parity mode must be disabled prior to gear-down mode entry. The two modes may
be enabled after tSYNC_GEAR and tCMD_GEAR periods have been satisfied. The gener-
al sequence for operation in 1/4 rate during initialization is as follows:
1. The device defaults to a 1N mode internal clock at power-up/reset.
2. Assertion of reset.
3. Assertion of CKE enables the DRAM.
4. MRS is accessed with a low-frequency N × tCK gear-down MRS command. (NtCK
static MRS command is qualified by 1N CS_n. )
5. The memory controller will send a 1N sync pulse with a low-frequency N × tCK
NOP command. tSYNC_GEAR is an even number of clocks. The sync pulse is on an
even edge clock boundary from the MRS command.
6. Initialization sequence, including the expiration of tDLLK and tZQinit, starts in 2N
mode after tCMD_GEAR from 1N sync pulse.
The device resets to 1N gear-down mode after entering self refresh. The general se-
quence for operation in gear-down after self refresh exit is as follows:
1. MRS is set to 1, via MR3[3], with a low-frequency N × tCK gear-down MRS com-
mand.
a. The NtCK static MRS command is qualified by 1N CS_n, which meets tXS or
tXS_ABORT.
b. Only a REFRESH command may be issued to the DRAM before the NtCK stat-
ic MRS command.
2. The DRAM controller sends a 1N sync pulse with a low-frequency N × tCK NOP
command.
a. tSYNC_GEAR is an even number of clocks.
b. The sync pulse is on even edge clock boundary from the MRS command.
3. A valid command not requiring locked DLL is available in 2N mode after
tCMD_GEAR from the 1N sync pulse.
a. A valid command requiring locked DLL is available in 2N mode after tXSDLL

or tDLLK from the 1N sync pulse.
4. If operation is in 1N mode after self refresh exit, N × tCK MRS command or sync
pulse is not required during self refresh exit. The minimum exit delay to the first
valid command is tXS, or tXS_ABORT.
The DRAM may be changed from 2N to 1N by entering self refresh mode, which will re-
set to 1N mode. Changing from 2N to by any other means can result in loss of data and
make operation of the DRAM uncertain.
When operating in 2N gear-down mode, the following MR settings apply:
• CAS latency (MR0[6:4,2]): Even number of clocks
• Write recovery and read to precharge (MR0[11:9]): Even number of clocks
• Additive latency (MR1[4:3]): CL - 2
• CAS WRITE latency (MR2 A[5:3]): Even number of clocks
Gear-Down Mode
• CS to command/address latency mode (MR4[8:6]): Even number of clocks

• CA parity latency mode (MR5[2:0]): Even number of clocks
Figure 40: Clock Mode Change from 1/2 Rate to 1/4 Rate (Initialization)
TdkN1 TdkN + Neven2
CK_c
CK_t
tCKSRX
DRAM
internal CLK
RESET_n
CKE
tXPR_GEAR tSYNC_GEAR tCMD_GEAR
1N sync pulse 2N mode
CS_n
tGEAR_setup tGEAR_hold tGEAR_setup tGEAR_hold
Command MRS NOP Valid
Configure DRAM 1) After tSYNC_GEAR from gear down command, internal clock rate is changed at TdkN
to 1/4 rate 2) After tSYNC_GEAR + tCMD_GEAR from gear down command, both internal clock rate and
Time Break Don’t Care command cycle are changed at TdkN+Neven
Figure 41: Clock Mode Change After Exiting Self Refresh

TdkN1 TdkN + Neven2
CK_c
CK_t
DRAM
internal CLK
CKE
tXPR_GEAR tSYNC_GEAR tCMD_GEAR
1N sync pulse 2N mode
CS_n
tGEAR_setup tGEAR_hold tGEAR_setup tGEAR_hold
Command MRS NOP Valid
Configure DRAM 1) After tSYNC_GEAR from gear down command, internal clock rate is changed at TdkN
to 1/4 rate 2) After tSYNC_GEAR + tCMD_GEAR from gear down command, both internal clock rate and
Time Break Don’t Care command cycle are changed at TdkN+Neven
CCMTD-1725822587-10418
Figure 42: Comparison Between Gear-Down Disable and Gear-Down Enable

CK_c
CK_t
AL = 0 (geardown = disable)
Command ACT DES DES DES DES READ DES DES DES DES DES DES DES DES DES DES DES DES
DQ DO DO DO DO DO DO DO DO
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
tRCD = 16 RL =CL= 16 (AL = 0)
AL = CL - 1 (geardown = disable) READ

Command ACT READ DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
RL = AL + CL = 31 (AL = CL - 1 = 15)
READ
Command ACT READ DES DES DES DES DES DES DES
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
AL + CL = RL = 30 (AL = CL - 2 = 14)
Time Break Transitioning Data Don’t Care

103

Gear-Down Mode
Maximum Power-Saving Mode

Maximum power-saving mode provides the lowest power mode where data retention is
not required. When the device is in the maximum power-saving mode, it does not
maintain data retention or respond to any external command, except the MAXIMUM
POWER SAVING MODE EXIT command and during the assertion of RESET_n signal
LOW. This mode is more like a “hibernate mode” than a typical power-saving mode.
The intent is to be able to park the DRAM at a very low-power state; the device can be
switched to an active state via the per-DRAM addressability (PDA) mode.
Maximum Power-Saving Mode Entry

Maximum power-saving mode is entered through an MRS command. For devices with
shared control/address signals, a single DRAM device can be entered into the maxi-
mum power-saving mode using the per-DRAM addressability MRS command. Large
CS_n hold time to CKE upon the mode exit could cause DRAM malfunction; as a result,
CA parity, CAL, and gear-down modes must be disabled prior to the maximum power-
saving mode entry MRS command.
The MRS command may use both address and DQ information, as defined in the Per-
DRAM Addressability section. As illustrated in the figure below, after tMPED from the
mode entry MRS command, the DRAM is not responsive to any input signals except
CKE, CS_n, and RESET_n. All other inputs are disabled (external input signals may be-
come High-Z). The system will provide a valid clock until tCKMPE expires, at which time
clock inputs (CK) should be disabled (external clock signals may become High-Z).
Figure 43: Maximum Power-Saving Mode Entry

Ta0 Ta1 Ta2 Tb0 Tb1 Tb3 Tc0 Tc1 Tc2 Tc3 Tc4 Tc5 Tc6 Tc7 Tc8 Tc9 Tc10 Tc11
CK_c
CK_t
tCKMPE
MR4[A1=1]
MPSM Enable)
Command DES MRS DES DES DES
tMPED
Address Valid
CS_n
CKE CKE LOW makes CS_n a care; CKE LOW followed by CS_n LOW followed by CKE HIGH exits mode
RESET_n
Maximum Power-Saving Mode Entry in PDA

The sequence and timing required for the maximum power-saving mode with the per-
DRAM addressability enabled is illustrated in the figure below.
Figure 44: Maximum Power-Saving Mode Entry with PDA

Ta0 Ta1 Ta2 Tb0 Tb1 Tb3 Tb4 Tb5 Tb6 Tb7 Tb8 Tb9 Tc0 Tc1 Tc2 Td0 Td1 Td2
CK_c
CK_t
MR4[A1 = 1]
MPSM Enable)
Command DES MRS DES DES DES DES DES DES DES DES DES DES DES DES
tCKMPE
CS_n
CKE
AL + CWL tMPED
DQS_t
DQS_c tPDA_S tPDA_H
DQ0
RESET_n
CKE Transition During Maximum Power-Saving Mode

The following figure shows how to maintain maximum power-saving mode even though
the CKE input may toggle. To prevent the device from exiting the mode, CS_n should be
HIGH at the CKE LOW-to-HIGH edge, with appropriate setup (tMPX_S) and hold
(tMPX_H) timings.
Figure 45: Maintaining Maximum Power-Saving Mode with CKE Transition
CLK
CMD
CS_n
tMPX_S tMPX_HH
CKE
RESET_n
Don’t Care
Maximum Power-Saving Mode Exit

To exit the maximum power-saving mode, CS_n should be LOW at the CKE LOW-to-
HIGH transition, with appropriate setup (tMPX_S) and hold (tMPX_LH) timings, as
shown in the figure below. Because the clock receivers (CK_t, CK_c) are disabled during
this mode, CS_n = LOW is captured by the rising edge of the CKE signal. If the CS_n sig-
nal level is detected LOW, the DRAM clears the maximum power-saving mode MRS bit
and begins the exit procedure from this mode. The external clock must be restarted and
be stable by tCKMPX before the device can exit the maximum power-saving mode. Dur-
ing the exit time (tXMP), only NOP and DES commands are allowed: NOP during
tMPX_LH and DES the remainder of tXMP. After tXMP expires, valid commands not re-
quiring a locked DLL are allowed; after tXMP_DLL expires, valid commands requiring a
locked DLL are allowed.
Figure 46: Maximum Power-Saving Mode Exit

Ta0 Ta1 Ta2 Ta3 Tb0 Tb1 Tb2 Tb3 Tc0 Tc1 Tc2 Tc4 Td0 Td1 Td2 Td3 Te0 Te1
CK_c
CK_t
tCKMPX
Command NOP NOP NOP NOP NOP DES DES DES DES Valid DES DES
tMPX_LH
CS_n
tMPX_S
CKE
tXMP
tXMP_DLL
RESET_n
Command/Address Parity
Command/address (CA) parity takes the CA parity signal (PAR) input carrying the parity
bit for the generated address and commands signals and matches it to the internally
generated parity from the captured address and commands signals. CA parity is suppor-
ted in the DLL enabled state only; if the DLL is disabled, CA parity is not supported.
Figure 47: Command/Address Parity Operation
DRAM Controller DRAM
CMD/ADDR CMD/ADDR
Even parity CMD/ADDR Even parity

GEN GEN
Even parity bit
Even parity bit

Compare
parity
bit
CA parity is disabled or enabled via an MRS command. If CA parity is enabled by pro-

gramming a non-zero value to CA parity latency in the MR, the DRAM will ensure that
there is no parity error before executing commands. There is an additional delay re-
quired for executing the commands versus when parity is disabled. The delay is pro-
grammed in the MR when CA parity is enabled (parity latency) and applied to all com-
mands which are registered by CS_n (rising edge of CK_t and falling CS_n). The com-
mand is held for the time of the parity latency (PL) before it is executed inside the de-
vice. The command captured by the input clock has an internal delay before executing
and is determined with PL. ALERT_n will go active when the DRAM detects a CA parity
error.
CA parity covers ACT_n, RAS_n/A16, CAS_n/A15, WE_n/A14, the address bus including
bank address and bank group bits, and C[2:0] on 3DS devices; the control signals CKE,
ODT, and CS_n are not covered. For example, for a 4Gb x4 monolithic device, parity is
computed across BG[1:0], BA[1:0], A16/RAS_n, A15/CAS_n, A14/ WE_n, A[13:0], and
ACT_n. The DRAM treats any unused address pins internally as zeros; for example, if a
common die has stacked pins but the device is used in a monolithic application, then
the address pins used for stacking and not connected are treated internally as zeros.
The convention for parity is even parity; for example, valid parity is defined as an even
number of ones across the inputs used for parity computation combined with the pari-
ty signal. In other words, the parity bit is chosen so that the total number of ones in the
transmitted signal, including the parity bit, is even.
If a DRAM device detects a CA parity error in any command qualified by CS_n, it will
perform the following steps:
1. Ignore the erroneous command. Commands in the MAX NnCK window
(tPAR_UNKNOWN) prior to the erroneous command are not guaranteed to be exe-
cuted. When a READ command in this NnCK window is not executed, the device
does not activate DQS outputs. If WRITE CRC is enabled and a WRITE CRC occurs
during the tPAR_UNKNOWN window, the WRITE CRC Error Status Bit located at
MR5[3] may or may not get set. When CA Parity and WRITE CRC are both enabled
and a CA Parity occurs, the WRITE CRC Error Status Bit should be reset.
2. Log the error by storing the erroneous command and address bits in the MPR er-
ror log.
3. Set the parity error status bit in the mode register to 1. The parity error status bit
must be set before the ALERT_n signal is released by the DRAM (that is,
tPAR_ALERT_ON + tPAR_ALERT_PW (MIN)).
4. Assert the ALERT_n signal to the host (ALERT_n is active LOW) within
tPAR_ALERT_ON time.
5. Wait for all in-progress commands to complete. These commands were received
tPAR_UNKOWN before the erroneous command.
6. Wait for tRAS (MIN) before closing all the open pages. The DRAM is not executing
any commands during the window defined by (tPAR_ALERT_ON +
tPAR_ALERT_PW).
7. After tPAR_ALERT_PW (MIN) has been satisfied, the device may de-assert
ALERT_n.
a. When the device is returned to a known precharged state, ALERT_n is al-
lowed to be de-asserted.
8. After (tPAR_ALERT_PW (MAX)) the DRAM is ready to accept commands for nor-
mal operation. Parity latency will be in effect; however, parity checking will not re-
sume until the memory controller has cleared the parity error status bit by writing
a zero. The DRAM will execute any erroneous commands until the bit is cleared;
unless persistent mode is enabled.
• It is possible that the device might have ignored a REFRESH command during
tPAR_ALERT_PW or the REFRESH command is the first erroneous frame, so it is rec-
ommended that extra REFRESH cycles be issued, as needed.

• The parity error status bit may be read anytime after tPAR_ALERT_ON +
tPAR_ALERT_PW to determine which DRAM had the error. The device maintains the
error log for the first erroneous command until the parity error status bit is reset to a
zero or a second CA parity occurs prior to resetting.
The mode register for the CA parity error is defined as follows: CA parity latency bits are
write only, the parity error status bit is read/write, and error logs are read-only bits. The
DRAM controller can only program the parity error status bit to zero. If the DRAM con-
troller illegally attempts to write a 1 to the parity error status bit, the DRAM can not be
certain that parity will be checked; the DRAM may opt to block the DRAM controller
from writing a 1 to the parity error status bit.
The device supports persistent parity error mode. This mode is enabled by setting
MR5[9] = 1; when enabled, CA parity resumes checking after the ALERT_n is de-asser-
ted, even if the parity error status bit remains a 1. If multiple errors occur before the er-
ror status bit is cleared the error log in MPR Page 1 should be treated as "Don’t Care." In
persistent parity error mode the ALERT_n pulse will be asserted and de-asserted by the
DRAM as defined with the MIN and MAX value tPAR_ALERT_PW. The DRAM controller
must issue DESELECT commands once it detects the ALERT_n signal, this response
time is defined as tPAR_ALERT_RSP. The following figures capture the flow of events on
the CA bus and the ALERT_n signal.
Table 36: Mode Register Setting for CA Parity
CA Parity Latency Erroneous CA

MR5[2:0]1 Applicable Speed Bin Parity Error Status Parity Persistent Mode Frame
000 = Disabled N/A
001 = 4 clocks 1600, 1866, 2133
010 = 5 clocks 2400, 2666 C[2:0], ACT_n, BG1,
BG0, BA[1:0], PAR,
011 = 6 clocks 2933, 3200 MR5 [4] 0 = Clear MR5 [9] 0 = DisabledMR5
A17, A16/RAS_n, A15/
100 = 8 clocks RFU MR5 [4] 1 = Error [9] 1 = Enabled
CAS_n, A14/WE_n,
101 = Reserved RFU A[13:0]
110 = Reserved RFU
111 = Reserved RFU
Notes: 1. Parity latency is applied to all commands.

2. Parity latency can be changed only from a CA parity disabled state; for example, a direct
change from PL = 3 to PL = 4 is not allowed. The correct sequence is PL = 3 to disabled to
PL = 4.
3. Parity latency is applied to WRITE and READ latency. WRITE latency = AL + CWL + PL.
READ latency = AL + CL + PL.
Figure 48: Command/Address Parity During Normal Operation

T0 T1 Ta0 Ta1 Ta2 Tb0 Tc0 Tc1 Td0 Te0 Te1
CK_c
CK_t
Command/
Valid 2 Valid 2 Valid 2 Error Valid Valid Valid DES2 DES2 Valid 3 Valid 3
Address
t > 2nCK tRP
tPAR_UNKNOWN 2 tPAR_ALERT_ON tPAR_ALERT_PW 1
ALERT_n
Valid 2 DES2 Command execution unknown

Error Valid Command not executed Don’t Care Time Break
Valid 3 Command executed
Notes: 1. DRAM is emptying queues. Precharge all and parity checking are off until parity error
status bit is cleared.
2. Command execution is unknown; the corresponding DRAM internal state change may
or may not occur. The DRAM controller should consider both cases and make sure that
the command sequence meets the specifications. If WRITE CRC is enabled and a WRITE
CRC occurs during the tPAR_UNKNOWN window, the WRITE CRC Error Status Bit located
at MR5[3] may or may not get set.
3. Normal operation with parity latency (CA parity persistent error mode disabled). Parity
checking is off until parity error status bit is cleared.
Figure 49: Persistent CA Parity Error Checking Operation

T0 T1 Ta0 Ta1 Ta2 Tb0 Tc0 Tc1 Td0 Te0 Te1
CK_c
CK_t
Command/
Valid 2 Valid 2 Valid 2 Error Valid Valid Valid DES DES DES Valid 3
Address
tPAR_ALERT_RSP t > 2nCK tRP
tPAR_UNKNOWN 2 tPAR_ALERT_ON tPAR_ALERT_PW 1
ALERT_n
Valid 2 DES Command execution unknown

Error Valid Command not executed
Valid 3 Command executed Don’t Care Time Break
Notes: 1. DRAM is emptying queues. Precharge all and parity check re-enable finished by
tPAR_ALERT_PW.
the command sequence meets the specifications. If WRITE CRC is enabled and a WRITE
CRC occurs during the tPAR_UNKNOWN window, the WRITE CRC Error Status Bit located
at MR5[3] may or may not get set
3. Normal operation with parity latency and parity checking (CA parity persistent error
mode enabled).
Figure 50: CA Parity Error Checking – SRE Attempt

T0 T1 Ta0 Ta1 Tb0 Tb1 Tc0 Tc1 Td0 Td1 Td2 Td3 Te0 Te1
CK_c
CK_t
tCPDED + PL tXP + PL
Command/
DES1, 5 Error2 DES1 DES6 DES6 DES5 Valid 3
Address
tIS tIS
CKE
t > 2nCK tRP
tIH
tPAR_ALERT_ON
Note 4 tPAR_ALERT_PW 1
ALERT_n
DES1, 5 DES6 DES5 Command execution unknown

Error2 DES1 Command not executed
Notes: 1. Only DESELECT command is allowed.

2. SELF REFRESH command error. The DRAM masks the intended SRE command and enters
precharge power-down.
checking is off until the parity error status bit cleared.
4. The controller cannot disable the clock until it has been capable of detecting a possible
CA parity error.
the command sequence meets the specifications.
6. Only a DESELECT command is allowed; CKE may go HIGH prior to Tc2 as long as DES
commands are issued.
Figure 51: CA Parity Error Checking – SRX Attempt

T0 Ta0 Ta1 Tb0 Tb1 Tc0 Tc1 Tc2 Td0 Td1 Te0 Tf0
CK_c
CK_t
Command/
SRX1 DES DES Error2 Valid 2 Valid 2 Valid 2 DES2, 3 DES2, 3 Valid 2, 4, 5 Valid 2, 4, 6 Valid 2, 4, 7
Address
t > 2nCK tRP
tIS
CKE
tPAR_UNKNOWN tPAR_ALERT_ON tPAR_ALERT_PW
ALERT_n
tXS_FAST 8
tXS
tXSDLL
SRX1 DES Valid 3, 5 Command execution unknown

Error Valid Command not executed Time Break Don’t Care
Valid 4,5,6,7 Command executed
Notes: 1. Self refresh abort = disable: MR4 [9] = 0.

2. Input commands are bounded by tXSDLL, tXS, tXS_ABORT, and tXS_FAST timing.
checking off until parity error status bit cleared.
5. Only an MRS (limited to those described in the SELF REFRESH Operation section), ZQCS,
or ZQCL command is allowed.
6. Valid commands not requiring a locked DLL.
7. Valid commands requiring a locked DLL.
8. This figure shows the case from which the error occurred after tXS_FAST. An error may
also occur after tXS_ABORT and tXS.
Figure 52: CA Parity Error Checking – PDE/PDX

T0 T1 Ta0 Ta1 Tb0 Tb1 Tc0 Tc1 Td0 Td1 Td2 Td3 Te0 Te1
CK_c
CK_t
tCPDED + PL tXP + PL
Command/ DES1 DES1

Error2 DES5 DES5 DES4 Valid 3
Address
tIS tIS
CKE
t > 2nCK tRP
tIH
tPAR_ALERT_ON tPAR_ALERT_PW 1
ALERT_n
DES4 DES5 Command execution unknown

Error2 DES1 Command not executed
Notes: 1. Only DESELECT command is allowed.

2. Error could be precharge or activate.
checking is off until parity error status bit cleared.
5. Only a DESELECT command is allowed; CKE may go HIGH prior to Td2 as long as DES
commands are issued.
Figure 53: Parity Entry Timing Example – tMRD_PAR

Ta0 Ta1 Ta2 Tb0 Tb1 Tb2
CK_c
CK_t
Command DES MRS DES DES MRS DES
Parity latency PL = 0 Updating setting PL = N
tMRD_PAR
Enable
parity
Note: 1. tMRD_PAR = tMOD + N; where N is the programmed parity latency.
Figure 54: Parity Entry Timing Example – tMOD_PAR

CK_c
CK_t
Command DES MRS DES DES Valid DES
Parity latency PL = 0 Updating setting PL = N
tMOD_PAR
Enable
parity
Note: 1. tMOD_PAR = tMOD + N; where N is the programmed parity latency.
Figure 55: Parity Exit Timing Example – tMRD_PAR

CK_c
CK_t
Command DES MRS DES DES MRS DES
Parity latency PL = N Updating setting
tMRD_PAR
Disable
parity
Note: 1. tMRD_PAR = tMOD + N; where N is the programmed parity latency.
Figure 56: Parity Exit Timing Example – tMOD_PAR

CK_c
CK_t
Command DES MRS DES DES Valid DES
Parity latency PL = N Updating setting
tMOD_PAR
Disable
parity
Note: 1. tMOD_PAR = tMOD + N; where N is the programmed parity latency.
CCMTD-1725822587-10418
Figure 57: CA Parity Flow Diagram

CA MR5[2:0] set parity latency (PL)
process start MR5[4] set parity error status to 0
MR5[9] enable/disable persistent mode
CA
latched in
Yes Persistent Yes Yes

CA parity CA parity
mode
enabled error
enabled
No No
No
Command
Good CA Ignore
MR5[4] = 0 Yes Yes execution
CA parity processed bad CMD
@ ADDR/CMD unknown
latched error
No
No
Command ALERT_n LOW MR5[4] = 0 Yes Log error/

Good CA Ignore @ ADDR/CMD
execution 44 to 144 CKs set parity status
processed bad CMD latched
unknown
114
Yes No
CA error

ALERT_n LOW Log error/ Internal
44 to 144 CKs set parity status precharge all
No
Good CA Bad CA
processed processed Internal
ALERT_n HIGH
precharge all
Normal
operation ready Operation ready?
Command
ALERT_n HIGH execution
unknown
Command Normal operation ready

execution MR5[4] reset to 0 if desired
unknown
Normal operation ready

MR5[4] reset to 0 if desired
DDR4 allows programmability of a single, specific DRAM on a rank. As an example, this
feature can be used to program different ODT or V REF values on each DRAM on a given
rank. Because per-DRAM addressability (PDA) mode may be used to program optimal
VREF for the DRAM, the data set up for first DQ0 transfer or the hold time for the last
DQ0 transfer cannot be guaranteed. The DRAM may sample DQ0 on either the first fall-
ing or second rising DQS transfer edge. This supports a common implementation be-
tween BC4 and BL8 modes on the DRAM. The DRAM controller is required to drive DQ0
to a stable LOW or HIGH state during the length of the data transfer for BC4 and BL8
cases. Note, both fixed and on-the-fly (OTF) modes are supported for BC4 and BL8 dur-
ing PDA mode.
1. Before entering PDA mode, write leveling is required.
• BL8 or BC4 may be used.
2. Before entering PDA mode, the following MR settings are possible:
• RTT(Park) MR5 A[8:6] = Enable
• RTT(NOM) MR1 A[10:8] = Enable
3. Enable PDA mode using MR3 [4] = 1. (The default programed value of MR3[4] = 0.)
4. In PDA mode, all MRS commands are qualified with DQ0. The device captures
DQ0 by using DQS signals. If the value on DQ0 is LOW, the DRAM executes the
MRS command. If the value on DQ0 is HIGH, the DRAM ignores the MRS com-
mand. The controller can choose to drive all the DQ bits.
5. Program the desired DRAM and mode registers using the MRS command and
DQ0.
6. In PDA mode, only MRS commands are allowed.
7. The MODE REGISTER SET command cycle time in PDA mode, AL + CWL + BL/2 -
0.5tCK + tMRD_PDA + PL, is required to complete the WRITE operation to the
mode register and is the minimum time required between two MRS commands.
8. Remove the device from PDA mode by setting MR3[4] = 0. (This command re-
quires DQ0 = 0.)
Note: Removing the device from PDA mode will require programming the entire MR3
when the MRS command is issued. This may impact some PDA values programmed
within a rank as the EXIT command is sent to the rank. To avoid such a case, the PDA
enable/disable control bit is located in a mode register that does not have any PDA
mode controls.
In PDA mode, the device captures DQ0 using DQS signals the same as in a normal
WRITE operation; however, dynamic ODT is not supported. Extra care is required for
the ODT setting. If RTT(NOM) MR1 [10:8] = enable, device data termination needs to be
controlled by the ODT pin, and applies the same timing parameters (defined below).
Symbol Parameter
DODTLon Direct ODT turnon latency
DODTLoff Direct ODT turn off latency
tADC RTT change timing skew
tAONAS Asynchronous RTT(NOM) turn-on delay
tAOFAS Asynchronous RTT(NOM) turn-off delay
Figure 58: PDA Operation Enabled, BL8
&.BF
&.BW
05$
3'$HQDEOH 056 056 056
W 02' &:/$/3/ W 05'B3'$
'46BW
'46BF
'4
W 3'$B6 W 3'$B+
'2'7/RII :/
2'7
'2'7/RQ :/
577 5773DUN 577120 5773DUN
Note: 1. RTT(Park) = Enable; RTT(NOM) = Enable; WRITE preamble set = 2tCK; and DLL = On.
Figure 59: PDA Operation Enabled, BC4
CK_c
CK_t
MR3 A4 = 1
(PDA enable) MRS MRS MRS
tMOD CWL+AL+PL tMRD_PDA
DQS_t
DQS_c
DQ0
tPDA_S tPDA_H
DODTLoff = WL-3
ODT
DODTLon = WL-3
RTT RTT(Park) RTT(NOM) RTT(Park)
Figure 60: MRS PDA Exit
&.BF
&.BW
05$
3'$GLVDEOH 056 9DOLG
&:/$/3/ W 02'B3'$
'46BW
'46BF
'4
W 3'$B6 W 3'$B+
'2'7/RII :/
2'7
'2'7/RQ :/
577 5773DUN 577120 5773DUN
VREFDQ Calibration
VREFDQ Calibration
The V REFDQ level, which is used by the DRAM DQ input receivers, is internally gener-
ated. The DRAM V REFDQ does not have a default value upon power-up and must be set
to the desired value, usually via V REFDQ calibration mode. If PDA or PPR modes (hPPR or
sPPR) are used prior to V REFDQ calibration, V REFDQ should initially be set at the midpoint
between the V DD,max, and the LOW as determined by the driver and ODT termination
selected with wide voltage swing on the input levels and setup and hold times of ap-
proximately 0.75UI. The memory controller is responsible for V REFDQ calibration to de-
termine the best internal V REFDQ level. The V REFDQ calibration is enabled/disabled via
MR6[7], MR6[6] selects Range 1 (60% to 92.5% of V DDQ) or Range 2 (45% to 77.5% of
VDDQ), and an MRS protocol using MR6[5:0] to adjust the V REFDQ level up and down.
MR6[6:0] bits can be altered using the MRS command if MR6[7] is enabled. The DRAM
controller will likely use a series of writes and reads in conjunction with V REFDQ adjust-
ments to obtain the best V REFDQ, which in turn optimizes the data eye.
The internal V REFDQ specification parameters are voltage range, step size, V REF step
time, V REF full step time, and V REF valid level. The voltage operating range specifies the
minimum required V REF setting range for DDR4 SDRAM devices. The minimum range is
defined by V REFDQ,min and V REFDQ,max. As noted, a calibration sequence, determined by
the DRAM controller, should be performed to adjust V REFDQ and optimize the timing
and voltage margin of the DRAM data input receivers. The internal V REFDQ voltage value
may not be exactly within the voltage range setting coupled with the V REF set tolerance;
the device must be calibrated to the correct internal V REFDQ voltage.
Figure 61: VREFDQ Voltage Range
VDDQ
VREF,max
VREF
range
VREF,min
VSWING small System variance

VSWING large Total range
VREFDQ Calibration
VREFDQ Range and Levels
Table 37: VREFDQ Range and Levels
MR6[5:0] Range 1 MR6[6] 0 Range 2 MR6[6] 1 MR6[5:0] Range 1 MR6[6] 0 Range 2 MR6[6] 1
00 0000 60.00% 45.00% 01 1010 76.90% 61.90%
00 0001 60.65% 45.65% 01 1011 77.55% 62.55%
00 0010 61.30% 46.30% 01 1100 78.20% 63.20%
00 0011 61.95% 46.95% 01 1101 78.85% 63.85%
00 0100 62.60% 47.60% 01 1110 79.50% 64.50%
00 0101 63.25% 48.25% 01 1111 80.15% 65.15%
00 0110 63.90% 48.90% 10 0000 80.80% 65.80%
00 0111 64.55% 49.55% 10 0001 81.45% 66.45%
00 1000 65.20% 50.20% 10 0010 82.10% 67.10%
00 1001 65.85% 50.85% 10 0011 82.75% 67.75%
00 1010 66.50% 51.50% 10 0100 83.40% 68.40%
00 1011 67.15% 52.15% 10 0101 84.05% 69.05%
00 1100 67.80% 52.80% 10 0110 84.70% 69.70%
00 1101 68.45% 53.45% 10 0111 85.35% 70.35%
00 1110 69.10% 54.10% 10 1000 86.00% 71.00%
00 1111 69.75% 54.75% 10 1001 86.65% 71.65%
01 0000 70.40% 55.40% 10 1010 87.30% 72.30%
01 0001 71.05% 56.05% 10 1011 87.95% 72.95%
01 0010 71.70% 56.70% 10 1100 88.60% 73.60%
01 0011 72.35% 57.35% 10 1101 89.25% 74.25%
01 0100 73.00% 58.00% 10 1110 89.90% 74.90%
01 0101 73.65% 58.65% 10 1111 90.55% 75.55%
01 0110 74.30% 59.30% 11 0000 91.20% 76.20%
01 0111 74.95% 59.95% 11 0001 91.85% 76.85%
01 1000 75.60% 60.60% 11 0010 92.50% 77.50%
01 1001 76.25% 61.25% 11 0011 to 11 1111 = Reserved
VREFDQ Step Size

The V REF step size is defined as the step size between adjacent steps. V REF step size rang-
es from 0.5% V DDQ to 0.8% V DDQ. However, for a given design, the device has one value
for V REF step size that falls within the range.
The V REF set tolerance is the variation in the V REF voltage from the ideal setting. This ac-
counts for accumulated error over multiple steps. There are two ranges for V REF set tol-
erance uncertainty. The range of V REF set tolerance uncertainty is a function of number
of steps n.
The V REF set tolerance is measured with respect to the ideal line, which is based on the
MIN and MAX V REF value endpoints for a specified range. The internal V REFDQ voltage
VREFDQ Calibration
value may not be exactly within the voltage range setting coupled with the V REF set tol-
erance; the device must be calibrated to the correct internal V REFDQ voltage.
Figure 62: Example of VREF Set Tolerance and Step Size
Actual VREF
output
Straight line
(endpoint fit)
VREF
VREF set
tolerance
VREF set
tolerance
VREF
step size
Digital Code
Note: 1. Maximum case shown.
VREFDQ Increment and Decrement Timing

The V REF increment/decrement step times are defined by V REF,time. V REF,time is defined
from t0 to t1, where t1 is referenced to the V REF voltage at the final DC level within the
VREF valid tolerance (VREF,val_tol). The V REF valid level is defined by V REF,val tolerance to
qualify the step time t1. This parameter is used to insure an adequate RC time constant
behavior of the voltage level change after any V REF increment/decrement adjustment.
VREFDQ Calibration
Figure 63: VREFDQ Timing Diagram for VREF,time Parameter
CK_c
CK_t
Command MRS
VREF setting
adjustment
DQ VREF Old VREF setting Updating VREF setting New VREF setting
VREF_time
t0 t1
Don’t Care
Note: 1. t0 is referenced to the MRS command clock

t1 is referenced to VREF,tol
VREFDQ calibration mode is entered via an MRS command, setting MR6[7] to 1 (0 disa-
bles V REFDQ calibration mode) and setting MR6[6] to either 0 or 1 to select the desired
range (MR6[5:0] are "Don't Care"). After V REFDQ calibration mode has been entered,
VREFDQ calibration mode legal commands may be issued once tVREFDQE has been sat-
isfied. Legal commands for V REFDQ calibration mode are ACT, WR, WRA, RD, RDA, PRE,
DES, and MRS to set V REFDQ values, and MRS to exit V REFDQ calibration mode. Also, after
VREFDQ calibration mode has been entered, “dummy” WRITE commands are allowed
prior to adjusting the V REFDQ value the first time V REFDQ calibration is performed after
initialization.
Setting V REFDQ values requires MR6[7] be set to 1 and MR6[6] be unchanged from the
initial range selection; MR6[5:0] may be set to the desired V REFDQ values. If MR6[7] is set
to 0, MR6[6:0] are not written. V REF,time-short or V REF,time-long must be satisfied after each
MR6 command to set V REFDQ value before the internal V REFDQ value is valid.
If PDA mode is used in conjunction with V REFDQ calibration, the PDA mode require-
ment that only MRS commands are allowed while PDA mode is enabled is not waived.
That is, the only V REFDQ calibration mode legal commands noted above that may be
used are the MRS commands: MRS to set V REFDQ values and MRS to exit V REFDQ calibra-
tion mode.
The last MR6[6:0] setting written to MR6 prior to exiting V REFDQ calibration mode is the
range and value used for the internal V REFDQ setting. V REFDQ calibration mode may be
exited when the DRAM is in idle state. After the MRS command to exit V REFDQ calibra-
tion mode has been issued, DES must be issued until tVREFDQX has been satisfied
where any legal command may then be issued. V REFDQ setting should be updated if the
die temperature changes too much from the calibration temperature.
The following are typical script when applying the above rules for V REFDQ calibration
routine when performing V REFDQ calibration in Range 1:
• MR6[7:6]10 [5:0]XXXXXXX.
VREFDQ Calibration
– Subsequent legal commands while in V REFDQ calibration mode: ACT, WR, WRA, RD,
RDA, PRE, DES, and MRS (to set V REFDQ values and exit V REFDQ calibration mode).
• All subsequent V REFDQ calibration MR setting commands are MR6[7:6]10
[5:0]VVVVVV.
– "VVVVVV" are desired settings for V REFDQ.
• Issue ACT/WR/RD looking for pass/fail to determine V CENT (midpoint) as needed.
• To exit V REFDQ calibration, the last two V REFDQ calibration MR commands are:
– MR6[7:6]10 [5:0]VVVVVV* where VVVVVV* = desired value for V REFDQ.
– MR6[7]0 [6:0]XXXXXXX to exit V REFDQ calibration mode.
The following are typical script when applying the above rules for V REFDQ calibration
routine when performing V REFDQ calibration in Range 2:
• MR6[7:6]11 [5:0]XXXXXXX.
– Subsequent legal commands while in V REFDQ calibration mode: ACT, WR, WRA, RD,
RDA, PRE, DES, and MRS (to set V REFDQ values and exit V REFDQ calibration mode).
• All subsequent V REFDQ calibration MR setting commands are MR6[7:6]11
[5:0]VVVVVV.
– "VVVVVV" are desired settings for V REFDQ.
• Issue ACT/WR/RD looking for pass/fail to determine V CENT (midpoint) as needed.
• To exit V REFDQ calibration, the last two V REFDQ calibration MR commands are:
– MR6[7:6]11 [5:0]VVVVVV* where VVVVVV* = desired value for V REFDQ.
– MR6[7]0 [6:0]XXXXXXX to exit V REFDQ calibration mode.
Note:
Range may only be set or changed when entering V REFDQ calibration mode; changing
range while in or exiting V REFDQ calibration mode is illegal.
Figure 64: VREFDQ Training Mode Entry and Exit Timing Diagram
T0 T1 Ta0 Ta1 Tb0 Tb1 Tc0 Tc1 Td0 Td1 Td2

CK_c
CK_t
Command DES MRS DES CMD DES CMD DES MRS1,2 DES WR DES
tVREFDQE tVREFDQX
VREFDQ training on New VREFDQ New VREFDQ VREFDQ training off

value or write value or write
Don’t Care
Notes: 1. New VREFDQ values are not allowed with an MRS command during calibration mode en-
try.
2. Depending on the step size of the latest programmed VREF value, VREF must be satisfied
before disabling VREFDQ training mode.
VREFDQ Calibration
Figure 65: VREF Step: Single Step Size Increment Case
VREF
Voltage
VREF
(VDDQ(DC))
VREF,val_tol
Step size
t1
Time
Figure 66: VREF Step: Single Step Size Decrement Case
VREF
Voltage
t1
Step size
VREF,val_tol
VREF
(VDDQ(DC))
Time
VREFDQ Calibration
Figure 67: VREF Full Step: From VREF,min to VREF,maxCase
VREF VREF
Voltage VREF,max (VDDQ(DC))
VREF,val_tol
Full range
t1
step
VREF,min
Time
Figure 68: VREF Full Step: From VREF,max to VREF,minCase
VREF,max
VREF
Voltage
Full range
step t1
VREF,val_tol
VREF,min
VREF
(VDDQ(DC))
Time
VREFDQ Target Settings

The V REFDQ initial settings are largely dependant on the ODT termination settings. The
table below shows all of the possible initial settings available for V REFDQ training; it is
unlikely the lower ODT settings would be used in most cases.
VREFDQ Calibration
Table 38: VREFDQ Settings (VDDQ = 1.2V)
RON ODT Vx – VIN LOW (mV) VREFDQ (mv) VREFDQ (%VDDQ)

34 ohm 600 900 75%
40 ohm 550 875 73%
48 ohm 500 850 71%
34 ohm 60 ohm 435 815 68%
80 ohm 360 780 65%
120 ohm 265 732 61%
240 ohm 150 675 56%
34 ohm 700 950 79%
40 ohm 655 925 77%
48 ohm 600 900 75%
48 ohm 60 ohm 535 865 72%
80 ohm 450 825 69%
120 ohm 345 770 64%
240 ohm 200 700 58%
Figure 69: VREFDQ Equivalent Circuit

VDDQ VDDQ
ODT
RXer
Vx
RON VREFDQ
(internal)
Connectivity Test Mode

Connectivity test (CT) mode is similar to boundary scan testing but is designed to sig-
nificantly speed up the testing of electrical continuity of pin interconnections between
the device and the memory controller on the PC boards. Designed to work seamlessly
with any boundary scan device, CT mode is supported in all ×4, ×8, and ×16 non-3DS
devices (JEDEC states CT mode for ×4 and ×8 is not required on 4Gb and is an optional
feature on 8Gb and above). 3DS devices do not support CT mode and the TEN pin
should be considered RFU maintained LOW at all times.
Contrary to other conventional shift-register-based test modes, where test patterns are
shifted in and out of the memory devices serially during each clock, the CT mode allows
test patterns to be entered on the test input pins in parallel and the test results to be
extracted from the test output pins of the device in parallel. These two functions are al-
so performed at the same time, significantly increasing the speed of the connectivity
check. When placed in CT mode, the device appears as an asynchronous device to the
external controlling agent. After the input test pattern is applied, the connectivity test
results are available for extraction in parallel at the test output pins after a fixed propa-
gation delay time.
Note: A reset of the device is required after exiting CT mode (see RESET and Initializa-
tion Procedure).
Pin Mapping
Only digital pins can be tested using the CT mode. For the purposes of a connectivity
check, all the pins used for digital logic in the device are classified as one of the follow-
ing types:
• Test enable (TEN): When asserted HIGH, this pin causes the device to enter CT mode.
In CT mode, the normal memory function inside the device is bypassed and the I/O
pins appear as a set of test input and output pins to the external controlling agent.
Additionally, the device will set the internal V REFDQ to V DDQ × 0.5 during CT mode
(this is the only time the DRAM takes direct control over setting the internal V REFDQ).
The TEN pin is dedicated to the connectivity check function and will not be used dur-
ing normal device operation.
• Chip select (CS_n): When asserted LOW, this pin enables the test output pins in the
device. When de-asserted, these output pins will be High-Z. The CS_n pin in the de-
vice serves as the CS_n pin in CT mode.
• Test input: A group of pins used during normal device operation designated as test
input pins. These pins are used to enter the test pattern in CT mode.
• Test output: A group of pins used during normal device operation designated as test
output pins. These pins are used for extraction of the connectivity test results in CT
mode.
• RESET_n: This pin must be fixed high level during CT mode, as in normal function.
Table 39: Connectivity Mode Pin Description and Switching Levels
CT Mode
Pins Pin Name During Normal Memory Operation Switching Level Notes
Test enable TEN CMOS (20%/80% VDD) 1, 2
Chip select CS_n VREFCA ±200mV 3
BA[1:0], BG[1:0], A[9:0], A10/AP, A11, A12/BC_n, A13, WE_n/A14, VREFCA ±200mV 3
A
CAS_n/A15, RAS_n/A16, CKE, ACT_n, ODT, CLK_t, CLK_c, PAR
Test B LDM_n/LDBI_n, UDM_n/UDBI_n; DM_n/DBI_n VREFDQ ±200mV 4
input
C ALERT_n CMOS (20%/80% VDD) 2, 5
D RESET_n CMOS (20%/80% VDD) 2
Test DQ[15:0], UDQS_t, UDQS_c, LDQS_t, LDQS_c; DQS_t, DQS_c VTT ±100mV 6
output
Notes: 1. TEN: Connectivity test mode is active when TEN is HIGH and inactive when TEN is LOW.
TEN must be LOW during normal operation.
2. CMOS is a rail-to-rail signal with DC HIGH at 80% and DC LOW at 20% of VDD (960mV
for DC HIGH and 240mV for DC LOW.)
3. VREFCA should be VDD/2.
4. VREFDQ should be VDDQ/2.
5. ALERT_n switching level is not a final setting.
6. VTT should be set to VDD/2.
Minimum Terms Definition for Logic Equations

The test input and output pins are related by the following equations, where INV de-
notes a logical inversion operation and XOR a logical exclusive OR operation:
MT0 = XOR (A1, A6, PAR)

MT1 = XOR (A8, ALERT_n, A9)
MT2 = XOR (A2, A5, A13) or XOR (A2, A5, A13, A17)
MT3 = XOR (A0, A7, A11)
MT4 = XOR (CK_c, ODT, CAS_n/A15)
MT5 = XOR (CKE, RAS_n/A16, A10/AP)
MT6 = XOR (ACT_n, A4, BA1)
MT7 = ×16: XOR (DMU_n/DBIU_n , DML_n/DBIL_n, CK_t)
= ×8: XOR (BG1, DML_n/DBIL_n, CK_t)
= ×4: XOR (BG1, CK_t)
MT8 = XOR (WE_n/A14, A12 / BC, BA0)
MT9 = XOR (BG0, A3, RESET_n and TEN)
Logic Equations for a ×4 Device
DQ0 = XOR (MT0, MT1)

DQS_t = MT8
DQS_c = MT9

DQ0 = MT0 DQ5 = MT5
DQ1 = MT1 DQ6 = MT6
DQ2 = MT2 DQ7 = MT7
DQ3 = MT3 DQS_t = MT8
DQ4 = MT4 DQS_c = MT9

DQ0 = MT0 DQ10 = INV DQ2
DQ6 = MT6 LDQS_t = MT8
DQ7 = MT7 LDQS_c = MT9
DQ8 = INV DQ0 UDQS_t = INV LDQS_t
DQ9 = INV DQ1 UDQS_c = INV LDQS_c
CT Input Timing Requirements

Prior to the assertion of the TEN pin, all voltage supplies, including V REFCA, must be val-
id and stable and RESET_n registered high prior to entering CT mode. Upon the asser-
tion of the TEN pin HIGH with RESET_n, CKE, and CS_n held HIGH; CLK_t, CLK_c, and
CKE signals become test inputs within tCTECT_Valid. The remaining CT inputs become
valid tCT_Enable after TEN goes HIGH when CS_n allows input to begin sampling, pro-
vided inputs were valid for at least tCT_Valid. While in CT mode, refresh activities in the
memory arrays are not allowed; they are initiated either externally (auto refresh) or in-
ternally (self refresh).
The TEN pin may be asserted after the DRAM has completed power-on. After the DRAM
is initialized and V REFDQ is calibrated, CT mode may no longer be used. The TEN pin
may be de-asserted at any time in CT mode. Upon exiting CT mode, the states and the
integrity of the original content of the memory array are unknown. A full reset of the
memory device is required.
After CT mode has been entered, the output signals will be stable within tCT_Valid after
the test inputs have been applied as long as TEN is maintained HIGH and CS_n is main-
tained LOW.
Figure 70: Connectivity Test Mode Entry

Ta Tb Tc Td
CK_t
Valid input Valid input
CK_c
tCKSRX tCT_IS
T = 10ns tIS tCT_IS
CKE Valid input Valid input

tCTCKE_Valid
T = 200μs T = 500μs
RESET_n
tCT_IS
TEN
tCTCKE_Valid>10ns
tCT_Enable
tCT_IS >0ns
CS_n
tCT_IS
CT Inputs Valid input Valid input

tCT_Valid tCT_Valid
tCT_Valid
CT Outputs Valid Valid
Don’t Care
Excessive Row Activation
Excessive Row Activation

Rows can be accessed a limited number of times within a certain time period before ad-
jacent rows require refresh. The maximum activate count (MAC) is the maximum num-
ber of activates that a single row can sustain within a time interval of equal to or less
than the maximum activate window (tMAW) before the adjacent rows need to be re-
freshed, regardless of how the activates are distributed over tMAW.
Micron's DDR4 devices automatically perform a type of TRR mode in the background
and provide an MPR Page 3 MPR3[3:0] of 1000, indicating there is no restriction to the
number of ACTIVATE commands to a given row in a refresh period provided DRAM
tim-ing specifications are not violated. However, specific attempts to by-pass TRR may
result in data disturb.
Table 40: MAC Encoding of MPR Page 3 MPR3

[7] [6] [5] [4] [3] [2] [1] [0] MAC Comments
x x x x 0 0 0 0 Untested The device has not been tested for MAC.
x x x x 0 0 0 1 tMAC = 700K
x x x x 0 0 1 0 tMAC = 600K
x x x x 0 0 1 1 tMAC = 500K
x x x x 0 1 0 0 tMAC = 400K
x x x x 0 1 0 1 tMAC = 300K
x x x x 0 1 1 0 Reserved
x x x x 0 1 1 1 tMAC = 200K
x x x x 1 0 0 0 Unlimited There is no restriction to the number of AC-
TIVATE commands to a given row in a re-
fresh period provided DRAM timing specifi-
cations are not violated.
x x x x : : : : Reserved
Note: 1. MAC encoding in MPR Page 3 MPR3.
Post Package Repair
Post Package Repair

Post Package Repair
JEDEC defines two modes of Post Package Repair (PPR): soft Post Package Repair (sPPR)
and hard Post Package Repair (hPPR). sPPR is non-persistent so the repair row maybe
altered; that is, sPPR is NOT a permanent repair and even though it will repair a row, the
repair can be reversed, reassigned via another sPPR, or made permanent via hPPR.
Hard Post Package Repair is persistent so once the repair row is assigned for a hPPR ad-
dress, further PPR commands to a previous hPPR section should not be performed, that
is, hPPR is a permanent repair; once repaired, it cannot be reversed. The controller pro-
vides the failing row address in the hPPR/sPPR sequence to the device to perform the
row repair. hPPR Mode and sPPR Mode may not be enabled at the same time.
JEDEC states hPPR is optional for 4Gb and sPPR is optional for 4Gb and 8Gb parts how-
ever Micron 4Gb and 8Gb DDR4 DRAMs should have both sPPR and hPPR support. The
hPPR support is identified via an MPR read from MPR Page 2, MPR0[7] and sPPR sup-
port is identified via an MPR read from MPR Page 2, MPR0[6].
The JEDEC minimum support requirement for DDR4 PPR (hPPR or sPPR) is to provide
one row of repair per bank group (BG), x4/x8 have 4 BG and x16 has 2 BG; this is a total
of 4 repair rows available on x4/x8 and 2 repair rows available on x16. Micron PPR sup-
port exceeds the JEDEC minimum requirements; Micron DDR4 DRAMs have at least
one row of repair for each bank which is essentially 4 row repairs per BG for a total of 16
repair rows for x4 and x8 and 8 repair rows for x16; a 4x increase in repair rows.
JEDEC requires the user to have all sPPR row repair addresses reset and cleared prior to
enabling hPPR Mode. Micron DDR4 PPR does not have this restriction, the existing
sPPR row repair addresses are not required to be cleared prior to entering hPPR mode.
Each bank in a BG is PPR independent: sPPR or hPPR issued to a bank will not alter a
sPPR row repair existing in a different bank.
sPPR followed by sPPR to same bank
When PPR is issued to a bank for the first time and is a sPPR command, the repair row
will be a sPPR. When a subsequent sPPR is issued to the same bank, the previous sPPR
repair row will be cleared and used for the subsequent sPPR address as the sPPR opera-
tion is non-persistent.
sPPR followed by hPPR to same bank
When a PPR is issued to a bank for the first time and is a sPPR command, the repair row
will be a sPPR. When a subsequent hPPR is issued to the same bank, the initial sPPR
repair row will be cleared and used for the hPPR address. If a further subsequent PPR
(hPPR or sPPR) is issued to the same bank, the further subsequent PPR ( hPPR or sPPR)
repair row will not clear or overwrite the previous hPPR address as the hPPR operation
is persistent.
hPPR followed by hPPR or sPPR to same bank
When a PPR is issued to a bank for the first time and is a hPPR command, the repair row
will be a hPPR. When a subsequent PPR (hPPR or sPPR) is issued to the same bank, the
subsequent PPR ( hPPR or sPPR) repair row will not clear or overwrite the initial hPPR
address as the initial hPPR is persistent.
Post Package Repair
Hard Post Package Repair

All banks must be precharged and idle. DBI and CRC modes must be disabled. Both
sPPR and hPPR must be disabled. sPPR is disabled with MR4[5] = 0. hPPR is disabled
with MR4[13] = 0, which is the normal state, and hPPR is enabled with MR4 [13]= 1,
which is the hPPR enabled state. There are two forms of hPPR mode. Both forms of
hPPR have the same entry requirement as defined in the sections below. The first com-
mand sequence uses a WRA command and supports data retention with a REFRESH
operation except for the bank containing the row that is being repaired; JEDEC has re-
laxed this requirement and allows BA[0] to be a don't care regarding the banks which
are not required to maintain data a REFRESH operation during hPPR. The second com-
mand sequence uses a WR command (a REFRESH operation can't be performed in this
command sequence). The second command sequence doesn't support data retention
for the target DRAM.
hPPR Row Repair - Entry

As stated above, all banks must be precharged and idle. DBI and CRC modes must be
disabled, and all timings must be followed as shown in the timing diagram that follows.
All other commands except those listed in the following sequences are illegal.
1. Issue MR4[13] 1 to enter hPPR mode enable.
a. All DQ are driven HIGH.
2. Issue four consecutive guard key commands (shown in the table below) to MR0
with each command separated by tMOD. The PPR guard key settings are the same
whether performing sPPR or hPPR mode.
a. Any interruption of the key sequence by other commands, such as ACT, WR,
RD, PRE, REF, ZQ, and NOP, are not allowed.
b. If the guard key bits are not entered in the required order or interrupted with
other MR commands, hPPR will not be enabled, and the programming cycle
will result in a NOP.
c. When the hPPR entry sequence is interrupted and followed by ACT and WR
commands, these commands will be conducted as normal DRAM com-
mands.
d. JEDEC allows A6:0 to be "Don't Care" on 4Gb and 8Gb devices from a suppli-
er perspective and the user should rely on vendor datasheet.
Table 41: PPR MR0 Guard Key Settings
MR0 BG1:0 BA1:0 A17:12 A11 A10 A9 A8 A7 A6:0

First guard key 0 0 xxxxxx 1 1 0 0 1 1111111
Second guard key 0 0 xxxxxx 0 1 1 1 1 1111111
Third Guard key 0 0 xxxxxx 1 0 1 1 1 1111111
Fourth guard key 0 0 xxxxxx 0 0 1 1 1 1111111
hPPR Row Repair – WRA Initiated (REF Commands Allowed)

1. Issue an ACT command with failing BG and BA with the row address to be re-
paired.
2. Issue a WRA command with BG and BA of failing row address.
a. The address must be at valid levels, but the address is "Don't Care."
3. All DQ of the target DRAM should be driven LOW for 4nCK (bit 0 through bit 7)
after WL (WL = CWL + AL + PL) in order for hPPR to initiate repair.
Post Package Repair
a. Repair will be initiated to the target DRAM only if all DQ during bit 0 through
bit 7 are LOW. The bank under repair does not get the REFRESH command
applied to it.
b. Repair will not be initiated to the target DRAM if any DQ during bit 0 through
bit 7 is HIGH.
1. JEDEC states: All DQs of target DRAM should be LOW for 4tCK. If HIGH
is driven to all DQs of a DRAM consecutively for equal to or longer than
2tCK, then DRAM does not conduct hPPR and retains data if REF com-
mand is properly issued; if all DQs are neither LOW for 4tCK nor HIGH
for equal to or longer than 2tCK, then hPPR mode execution is un-
known.
c. DQS should function normally.
4. REF command may be issued anytime after the WRA command followed by WL +
4nCK + tWR + tRP.
a. Multiple REF commands are issued at a rate of tREFI or tREFI/2, however
back-to-back REF commands must be separated by at least tREFI/4 when the
DRAM is in hPPR mode.
b. All banks except the bank under repair will perform refresh.
5. Issue PRE after tPGM time so that the device can repair the target row during tPGM
time.
a. Wait tPGM_Exit after PRE to allow the device to recognize the repaired target
row address.
6. Issue MR4[13] 0 command to hPPR mode disable.
a. Wait tPGMPST for hPPR mode exit to complete.
b. After tPGMPST has expired, any valid command may be issued.
The entire sequence from hPPR mode enable through hPPR mode disable may be re-
peated if more than one repair is to be done.
After completing hPPR mode, MR0 must be re-programmed to a prehPPR mode state if
the device is to be accessed.
After hPPR mode has been exited, the DRAM controller can confirm if the target row
was repaired correctly by writing data into the target row and reading it back.
Figure 71: hPPR WRA – Entry

7 7 7D 7D 7E 7E 7F 7F 7G 7G 7H 7I 7J
&.BF
&.BW
&0' 056 '(6 056 '(6 056 '(6 056 '(6 056 '(6 $&7 :5$ '(6
%* 9DOLG 1$ 1$ 1$ 1$ 1$ %*I %*I 1$
%$ 9DOLG 1$ 1$ 1$ 1$ 1$ %$I %$I 1$
9DOLG
$''5 $ 1$ VW.H\ 1$ QG .H\ 1$ UG .H\ 1$ WK .H\ 1$ 9DOLG 9DOLG 1$
&.(
'46BW
'46BF
'4V
$OO%DQNV W 02' W 02' W 02' W 02' W 02' W 5&'

3UHFKDUJHG
DQGLGOHVWDWH
1RUPDO K335(QWU\ VW*XDUG.H\9DOLGDWH QG *XDUG.H\9DOLGDWH UG *XDUG.H\9DOLGDWH WK *XDUG.H\9DOLGDWH K3355HSDLU
0RGH
'RQ¶W&DUH
Post Package Repair
Figure 72: hPPR WRA – Repair and Exit

Te0 Tf0 Tg0 Tg1 Th0 Th1 Tj0 Tj1 Tj2 Tk0 Tk1 Tm0 Tm1 Tn0
CK_c
CK_t
CMD ACT WRA DES DES DES DES DES REF/DES REF/DES PRE REF/DES MRSx DES Valid
BG BGf BGf N/A N/A N/A N/A N/A N/A N/A Valid N/A Valid N/A Valid
BA BAf BAf N/A N/A N/A N/A N/A N/A N/A Valid N/A Valid N/A Valid
Valid Valid N/A N/A N/A N/A N/A N/A N/A Valid N/A Valid N/A Valid
ADDR (A13 = 0)
CKE
WL = CWL+AL+PL 4nCK tWR +tRP + 1nCK
DQS_t
DQS_c
DQs1
bit 0 bit 1 bit 6 bit 7
tRCD tPGM tPGM_Exit tPGMPST
All Banks
Precharged
and idle state
hPPR Recognition hPPR Exit Normal

hPPR Repair hPPR Repair hPPR Repair mode
Don’t Care
hPPR Row Repair – WR Initiated (REF Commands NOT Allowed)

1. Issue an ACT command with failing BG and BA with the row address to be re-
paired.
2. Issue a WR command with BG and BA of failing row address.
a. The address must be at valid levels, but the address is "Don't Care."
after WL (WL = CWL + AL + PL) in order for hPPR to initiate repair.
bit 7 are LOW.
bit 7 is HIGH.
2tCK, then DRAM does not conduct hPPR and retains data if REF com-
mand is properly issued; if all DQs are neither LOW for 4tCK nor HIGH
for equal to or longer than 2tCK, then hPPR mode execution is un-
known.
4. REF commands may NOT be issued at anytime while in PPT mode.
5. Issue PRE after tPGM time so that the device can repair the target row during tPGM
time.
a. Wait tPGM_Exit after PRE to allow the device to recognize the repaired target
row address.
6. Issue MR4[13] 0 command to hPPR mode disable.
a. Wait tPGMPST for hPPR mode exit to complete.
b. After tPGMPST has expired, any valid command may be issued.
The entire sequence from hPPR mode enable through hPPR mode disable may be re-
peated if more than one repair is to be done.
After completing hPPR mode, MR0 must be re-programmed to a prehPPR mode state if
the device is to be accessed.
After hPPR mode has been exited, the DRAM controller can confirm if the target row
Post Package Repair
Figure 73: hPPR WR – Entry

7 7 7D 7D 7E 7E 7F 7F 7G 7G 7H 7I 7J
&.BF
&.BW
&0' 056 '(6 056 '(6 056 '(6 056 '(6 056 '(6 $&7 :5 '(6
%* 9DOLG 1$ 1$ 1$ 1$ 1$ %*I %*I 1$
%$ 9DOLG 1$ 1$ 1$ 1$ 1$ %$I %$I 1$
9DOLG
&.(
:/ &:/
'46BW
'46BF
'4V
$OO%DQNV W 02' W 02' W 02' W 02' W 02' W 5&'

3UHFKDUJHG
DQGLGOHVWDWH
1RUPDO K335(QWU\ VW*XDUG.H\9DOLGDWH QG *XDUG.H\9DOLGDWH UG *XDUG.H\9DOLGDWH WK *XDUG.H\9DOLGDWH K3355HSDLU
0RGH
'RQ¶W&DUH
Figure 74: hPPR WR – Repair and Exit

CK_c
CK_t
CMD ACT WR DES DES DES DES DES DES DES PRE DES MRSx DES Valid
ADDR (A13 = 0)
CKE
WL = CWL + AL + PL 4nCK
DQS_t
DQS_c
DQs1
tRCD tPGM tPGM_Exit tPGMPST
All Banks
Precharged
and idle state
hPPR Recognition hPPR Exit Normal

hPPR Repair hPPR Repair hPPR Repair mode
Don’t Care
Table 42: DDR4 hPPR Timing Parameters DDR4-1600 through DDR4-3200
Parameter Symbol Min Max Unit

hPPR programming time tPGM ×4, ×8 1000 – ms
×16 2000 – ms
hPPR precharge exit time tPGM_Exit 15 – ns
hPPR exit time tPGMPST 50 – μs
sPPR Row Repair

Soft post package repair (sPPR) is a way to quickly, but temporarily, repair a row ele-
ment in a bank on a DRAM device, where hPPR takes longer but permanently repairs a
row element. sPPR mode is entered in a similar fashion as hPPR, sPPR uses MR4[5]
while hPPR uses MR4[13]. sPPR is disabled with MR4[5] = 0, which is the normal state,
and sPPR is enabled with MR4[5] = 1, which is the sPPR enabled state.
Post Package Repair
sPPR requires the same guard key sequence as hPPR to qualify the MR4 PPR entry. After
sPPR entry, an ACT command will capture the target bank and target row, herein seed
row, where the row repair will be made. After tRCD time, a WR command is used to se-
lect the individual DRAM, through the DQ bits, to transfer the repair address into an in-
ternal register in the DRAM. After a write recovery time and PRE command, the sPPR
mode can be exited and normal operation can resume.
The DRAM will retain the soft repair information as long as V DD remains within the op-
erating region unless rewritten by a subsequent sPPR entry to the same bank. If DRAM
power is removed or the DRAM is reset, the soft repair will revert to the unrepaired
state. hPPR and sPPR should not be enabled at the same time; Micron sPPR does not
have to be disabled and cleared prior to entering hPPR mode.
With sPPR, Micron DDR4 can repair one row per bank. When a subsequent sPPR re-
quest is made to the same bank, the subsequently issued sPPR address will replace the
previous sPPR address. When the hPPR resource for a bank is used up, the bank should
be assumed to not have available resources for sPPR. If a repair sequence is issued to a
bank with no repair resource available, the DRAM will ignore the programming se-
quence.
The bank receiving sPPR change is expected to retain memory array data in all rows ex-
cept for the seed row and its associated row addresses. If the data in the memory array
in the bank under sPPR repair is not required to be retained, then the handling of the
seed row’s associated row addresses is not of interest and can be ignored. If the data in
the memory array is required to be retained in the bank under sPPR mode, then prior to
executing the sPPR mode, the seed row and its associated row addresses should be
backed up and subsequently restored after sPPR has been completed. sPPR associated
seed row addresses are specified in the Table below; BA0 is not required by Micron
DRAMs however it is JEDEC reserved.
Table 43: sPPR Associated Rows
sPPR Associated Row Address

BA0* A17 A16 A15 A14 A13 A1 A0
All banks must be precharged and idle. DBI and CRC modes must be disabled, and all
sPPR timings must be followed as shown in the timing diagram that follows.
All other commands except those listed in the following sequences are illegal.
1. Issue MR4[5] 1 to enter sPPR mode enable.
a. All DQ are driven HIGH.
2. Issue four consecutive guard key commands (shown in the table below) to MR0
with each command separated by tMOD. Please note that JEDEC recently added
the four guard key entry used for hPPR to sPPR entry; early DRAMs may not re-
quire four guard key entry code. A prudent controller design should accommodate
either option in case an earlier DRAM is used.
a. Any interruption of the key sequence by other commands, such as ACT, WR,
RD, PRE, REF, ZQ, and NOP, are not allowed.
b. If the guard key bits are not entered in the required order or interrupted with
other MR commands, sPPR will not be enabled, and the programming cycle
will result in a NOP.
c. When the sPPR entry sequence is interrupted and followed by ACT and WR
commands, these commands will be conducted as normal DRAM com-
mands.
Post Package Repair
d. JEDEC allows A6:0 to be "Don't Care" on 4Gb and 8Gb devices from a suppli-
er perspective and the user should rely on vendor datasheet.
Table 44: PPR MR0 Guard Key Settings
MR0 BG1:0 BA1:0 A17:12 A11 A10 A9 A8 A7 A6:0

First guard key 0 0 xxxxxx 1 1 0 0 1 1111111
Second guard key 0 0 xxxxxx 0 1 1 1 1 1111111
Third guard key 0 0 xxxxxx 1 0 1 1 1 1111111
Fourth guard key 0 0 xxxxxx 0 0 1 1 1 1111111
3. After tMOD, issue an ACT command with failing BG and BA with the row address
to be repaired.
4. After tRCD, issue a WR command with BG and BA of failing row address.
a. The address must be at valid levels, but the address is a "Don't Care."
after WL (WL = CWL + AL + PL) in order for sPPR to initiate repair.
bit 7 are LOW.
bit 7 is HIGH.
the first 2tCK, then DRAM does not conduct hPPR and retains data if
REF command is properly issued; if all DQs are neither LOW for 4tCK
nor HIGH for equal to or longer than the first 2tCK, then hPPR mode ex-
ecution is unknown.
6. REF command may NOT be issued at anytime while in sPPR mode.
7. Issue PRE after tWR time so that the device can repair the target row during tWR
time.
a. Wait tPGM_Exit_s after PRE to allow the device to recognize the repaired tar-
get row address.
8. Issue MR4[5] 0 command to sPPR mode disable.
a. Wait tPGMPST_s for sPPR mode exit to complete.
b. After tPGMPST_s has expired, any valid command may be issued.
The entire sequence from sPPR mode enable through sPPR mode disable may be repea-
ted if more than one repair is to be done.
After sPPR mode has been exited, the DRAM controller can confirm if the target row
Post Package Repair
Figure 75: sPPR – Entry

7 7 7D 7D 7E 7E 7F 7F 7G 7G 7H 7I 7J
&.BF
&.BW
&0' 056 '(6 056 '(6 056 '(6 056 '(6 056 '(6 $&7 :5 '(6
%* 9DOLG 1$ 1$ 1$ 1$ 1$ %*I %*I 1$
%$ 9DOLG 1$ 1$ 1$ 1$ 1$ %$I %$I 1$
9DOLG
&.(
'46BW
'46BF
'4V
$OO%DQNV W 02' W 02' W 02' W 02' W 02' W 5&'

3UHFKDUJHG
DQGLGOHVWDWH
1RUPDO V335(QWU\ VW*XDUG.H\9DOLGDWH QG *XDUG.H\9DOLGDWH UG *XDUG.H\9DOLGDWH WK *XDUG.H\9DOLGDWH V3355HSDLU
0RGH
'RQ¶W&DUH
Figure 76: sPPR – Repair, and Exit

CK_c
CK_t
CMD ACT WR DES DES DES DES DES DES DES PRE DES MRS4 DES Valid
ADDR (A5=0)
CKE
WL = CWL + AL + PL 4nCK tWR
DQS_t
DQS_c
DQs1
tRCD tPGM_s tPGM_Exit_s tPGMPST_s
All Banks
Precharged
and idle state
sPPR Recognition sPPR Exit Normal

sPPR Repair sPPR Repair sPPR Repair sPPR Repair Mode
Don’t Care
Table 45: DDR4 sPPR Timing Parameters DDR4-1600 through DDR4-3200
Parameter Symbol Min Max Unit

sPPR programming time tPGM_s t RCD(MIN)+ WL + 4nCK – ns
+ tWR(MIN)
sPPR precharge exit time tPGM_Exit_s 20 – ns
sPPR exit time tPGMPST_s tMOD – ns
hPPR/sPPR Support Identifier
Table 46: DDR4 Repair Mode Support Identifier
MPR Page 2 A7 A6 A5 A4 A3 A2 A1 A0
UI0 UI1 UI2 UI3 UI4 UI5 UI6 UI7
ACTIVATE Command
Table 46: DDR4 Repair Mode Support Identifier (Continued)
MPR0 hPPR1 sPPR2 RTT_WR Temp sensor CRC RTT_WR
Notes: 1. 0 = hPPR mode is not available, 1 = hPPR mode is available.

2. 0 = sPPR mode is not available, 1 = sPPR mode is available.
3. Gray shaded areas are for reference only.
ACTIVATE Command
The ACTIVATE command is used to open (activate) a row in a particular bank for subse-
quent access. The values on the BG[1:0] inputs select the bank group, the BA[1:0] inputs
select the bank within the bank group, and the address provided on inputs A[17:0] se-
lects the row within the bank. This row remains active (open) for accesses until a PRE-
CHARGE command is issued to that bank. A PRECHARGE command must be issued be-
fore opening a different row in the same bank. Bank-to-bank command timing for AC-
TIVATE commands uses two different timing parameters, depending on whether the
banks are in the same or different bank group. tRRD_S (short) is used for timing be-
tween banks located in different bank groups. tRRD_L (long) is used for timing between
banks located in the same bank group. Another timing restriction for consecutive ACTI-
VATE commands [issued at tRRD (MIN)] is tFAW (fifth activate window). Because there
is a maximum of four banks in a bank group, the tFAW parameter applies across differ-
ent bank groups (five ACTIVATE commands issued at tRRD_L (MIN) to the same bank
group would be limited by tRC).
Figure 77: tRRD Timing

T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
CK_c
CK_t
Command ACT DES DES DES ACT DES DES DES DES DES ACT DES
tRRD_S tRRD_L
Bank
Group BG a BG b BG b
(BG)
Bank Bank c Bank c Bank d
Address Row n Row n Row n
Don’t Care
Notes: 1. tRRD_S; ACTIVATE-to-ACTIVATE command period (short); applies to consecutive ACTI-

VATE commands to different bank groups (that is, T0 and T4).
2. tRRD_L; ACTIVATE-to-ACTIVATE command period (long); applies to consecutive ACTI-
VATE commands to the different banks in the same bank group (that is, T4 and T10).
PRECHARGE Command
Figure 78: tFAW Timing

CK_c T0 Ta0 Tb0 Tc0 Tc1 Tc2 Td0 Td1
CK_t
Command ACT Valid ACT Valid ACT Valid ACT Valid Valid Valid ACT NOP
tRRD tRRD tRRD

tFAW
Bank
Group Valid Valid Valid Valid Valid
(BG)
Bank Valid Valid Valid Valid Valid
Address Valid Valid Valid Valid Valid
Don’t Care Time Break
Note: 1. tFAW; four activate windows.
PRECHARGE Command
The PRECHARGE command is used to deactivate the open row in a particular bank or
the open row in all banks. The bank(s) will be available for a subsequent row activation
for a specified time (tRP) after the PRECHARGE command is issued. An exception to
this is the case of concurrent auto precharge, where a READ or WRITE command to a
different bank is allowed as long as it does not interrupt the data transfer in the current
bank and does not violate any other timing parameters.
After a bank is precharged, it is in the idle state and must be activated prior to any READ
or WRITE commands being issued to that bank. A PRECHARGE command is allowed if
there is no open row in that bank (idle state) or if the previously open row is already in
the process of precharging. However, the precharge period will be determined by the
last PRECHARGE command issued to the bank.
The auto precharge feature is engaged when a READ or WRITE command is issued with
A10 HIGH. The auto precharge feature uses the RAS lockout circuit to internally delay
the PRECHARGE operation until the ARRAY RESTORE operation has completed. The
RAS lockout circuit feature allows the PRECHARGE operation to be partially or com-
pletely hidden during burst READ cycles when the auto precharge feature is engaged.
The PRECHARGE operation will not begin until after the last data of the burst write se-
quence is properly stored in the memory array.
REFRESH Command
The REFRESH command (REF) is used during normal operation of the device. This
command is nonpersistent, so it must be issued each time a refresh is required. The de-
vice requires REFRESH cycles at an average periodic interval of tREFI. When CS_n,
RAS_n/A16, and CAS_n/A15 are held LOW and WE_n/A14 HIGH at the rising edge of
the clock, the device enters a REFRESH cycle. All banks of the SDRAM must be pre-
charged and idle for a minimum of the precharge time, tRP (MIN), before the REFRESH
command can be applied. The refresh addressing is generated by the internal DRAM re-
fresh controller. This makes the address bits “Don’t Care” during a REFRESH command.
An internal address counter supplies the addresses during the REFRESH cycle. No con-
trol of the external address bus is required once this cycle has started. When the RE-
FRESH cycle has completed, all banks of the SDRAM will be in the precharged (idle)
REFRESH Command
state. A delay between the REFRESH command and the next valid command, except
DES, must be greater than or equal to the minimum REFRESH cycle time tRFC (MIN),
as shown in Figure 79 (page 141).
Note: The tRFC timing parameter depends on memory density.
In general, a REFRESH command needs to be issued to the device regularly every tREFI
interval. To allow for improved efficiency in scheduling and switching between tasks,
some flexibility in the absolute refresh interval is provided for postponing and pulling-
in the REFRESH command. A limited number REFRESH commands can be postponed
depending on refresh mode: a maximum of 8 REFRESH commands can be postponed
when the device is in 1X refresh mode; a maximum of 16 REFRESH commands can be
postponed when the device is in 2X refresh mode; and a maximum of 32 REFRESH
commands can be postponed when the device is in 4X refresh mode.
When 8 consecutive REFRESH commands are postponed, the resulting maximum inter-
val between the surrounding REFRESH commands is limited to 9 × tREFI (see Figure 80
(page 142)). For both the 2X and 4X refresh modes, the maximum interval between sur-
rounding REFRESH commands allowed is limited to 17 × tREFI2 and 33 × tREFI4, re-
spectively.
A limited number REFRESH commands can be pulled-in as well. A maximum of 8 addi-
tional REFRESH commands can be issued in advance or “pulled-in” in 1X refresh mode,
a maximum of 16 additional REFRESH commands can be issued when in advance in 2X
refresh mode, and a maximum of 32 additional REFRESH commands can be issued in
advance when in 4X refresh mode. Each of these REFRESH commands reduces the
number of regular REFRESH commands required later by one. Note that pulling in
more than the maximum allowed REFRESH commands in advance does not further re-
duce the number of regular REFRESH commands required later, so that the resulting
maximum interval between two surrounding REFRESH commands is limited to 9 × tRE-
FI (Figure 81 (page 142)), 17 × tRFEI2, or 33 × tREFI4. At any given time, a maximum of
16 additional REF commands can be issued within 2 × tREFI, 32 additional REF2 com-
mands can be issued within 4 × tREFI2, and 64 additional REF4 commands can be is-
sued within 8 × tREFI4 (larger densities are limited by tRFC1, tRFC2, and tRFC4, respec-
tively, which must still be met).
Figure 79: REFRESH Command Timing

T0 T1 Ta0 Ta1 Tb0 Tb1 Tb2 Tb3 Tc0 Tc1 Tc2 Tc3
CK_c
CK_t
Command REF DES DES REF DES DES Valid Valid Valid Valid Valid REF Valid Valid Valid
tRFC tRFC (MIN)
tREFI (MAX 9 × tREFI)
DRAM must be idle DRAM must be idle
Notes: 1. Only DES commands are allowed after a REFRESH command is registered until tRFC
(MIN) expires.
2. Time interval between two REFRESH commands may be extended to a maximum of 9 ×
tREFI.
REFRESH Command
Figure 80: Postponing REFRESH Commands (Example)

tREFI
9 × tREFI
W
tRFC
8 REF-Commands postponed
Figure 81: Pulling In REFRESH Commands (Example)

tREFI 9 × tREFI
W
tRFC
8 REF-Commands pulled-in
Temperature-Controlled Refresh Mode

During normal operation, temperature-controlled refresh (TCR) mode disabled, the de-
vice must have a REFRESH command issued once every tREFI, except for what is al-
lowed by posting (see REFRESH Command section). This means a REFRESH command
must be issued once every 0.975μs if T C is greater than 105°C, once every 1.95μs if T C is
greater than 95°C, once every 3.9μs if T C is greater than 85°C, and once every 7.8μs if T C
is less than or equal to 85°C, regardless of which Temperature Mode is selected
(MR4[2]). TCR mode is disabled by setting MR4[3] = 0 while TCR mode is enabled by
setting MR4[3] = 1. When TCR mode is enabled (MR4[3] = 1), the Temperature Mode
must be selected where MR4[2] = 0 enables the Normal Temperature Mode while
MR4[2] = 1 enables the Extended Temperature Mode.
When TCR mode is enabled, the device will register the externally supplied REFRESH
command and adjust the internal refresh period to be longer than tREFI of the normal
temperature range, when allowed, by skipping REFRESH commands with the proper
gear ratio. TCR mode has two Temperature Modes to select between the normal tem-
perature range and the extended temperature range; the correct Temperature Mode
must be selected so the internal control operates correctly. The DRAM must have the
correct refresh rate applied externally; the internal refresh rate is determined by the
DRAM based upon the temperature.
Normal Temperature Mode

REFRESH commands should be issued to the device with the refresh period equal to
tREFI of normal temperature range (0°C to 85°C). The system must guarantee that the
TC does not exceed 85°C. The device may adjust the internal refresh period to be longer
than tREFI of the normal temperature range by skipping external REFRESH commands
with the proper gear ratio when T C is below 85°C. The internal refresh period is auto-
matically adjusted inside the DRAM, and the DRAM controller does not need to provide
any additional control.
Extended Temperature Mode

REFRESH commands should be issued to the device with the refresh period equal to
tREFI of extended temperature range (85°C to 125°C). The system must guarantee that
the T C does not exceed 125°C. Even though the external refresh supports the extended
temperature range, the device may adjust its internal refresh period to be equal to or
longer than tREFI of the normal temperature range (0°C to 85°C) by skipping external
REFRESH commands with the proper gear ratio when T C is equal to or below 85°C. The
internal refresh period is automatically adjusted inside the DRAM, and the DRAM con-
troller does not need to provide any additional control.
Table 47: Normal tREFI Refresh (TCR Enabled)
Normal Temperature Mode Extended Temperature Mode

External Refresh Internal Refresh External Refresh Internal Refresh
Temperature Period Period Period Period
TC ื 85°C 7.8μs ุ7.8μs ุ7.8μs
3.9μs1
85°C < TC ื 95°C N/A 3.9μs
95°C < TC ื105°C N/A 1.95μs 1.95μs
Table 47: Normal tREFI Refresh (TCR Enabled) (Continued)
Normal Temperature Mode Extended Temperature Mode

External Refresh Internal Refresh External Refresh Internal Refresh
Temperature Period Period Period Period
105°C < TC ื 125°C N/A 0.975μs 0.975μs
Note: 1. If the external refresh period is slower than 3.9μs, the device will refresh internally at
too slow of a refresh rate and will violate refresh specifications.
Figure 82: TCR Mode Example1
Controller 85°C TC 95°C TC 85°C
REFRESH REFRESH REFRESH

External Internal
tREFI tREFI
3.9μs 3.9μs Internal
REFRESH REFRESH tREFI
7.8μs
REFRESH REFRESH
REFRESH REFRESH
Controller issues REFRESH External REFRESH At least every other

commands at extended commands are not external REFRESH
temperature rate ignored ignored
Note: 1. TCR enabled with Extended Temperature Mode selected.

Mode Register and Command Truth Table
The REFRESH cycle time (tRFC) and the average refresh interval (tREFI) can be pro-
grammed by the MRS command. The appropriate setting in the mode register will set a
single set of REFRESH cycle times and average refresh interval for the device (fixed
mode), or allow the dynamic selection of one of two sets of REFRESH cycle times and
average refresh interval for the device (on-the-fly mode [OTF]). OTF mode must be ena-
bled by MRS before any OTF REFRESH command can be issued.
Table 48: MRS Definition
MR3[8] MR3[7] MR3[6] Refresh Rate Mode

0 0 0 Normal mode (fixed 1x)
0 0 1 Fixed 2x
0 1 0 Fixed 4x
0 1 1 Reserved
1 0 0 Reserved
1 0 1 On-the-fly 1x/2x
1 1 0 On-the-fly 1x/4x
1 1 1 Reserved
There are two types of OTF modes (1x/2x and 1x/4x modes) that are selectable by pro-
gramming the appropriate values into the mode register. When either of the two OTF
modes is selected, the device evaluates the BG0 bit when a REFRESH command is is-
sued, and depending on the status of BG0, it dynamically switches its internal refresh
configuration between 1x and 2x (or 1x and 4x) modes, and then executes the corre-
sponding REFRESH operation.
Table 49: REFRESH Command Truth Table
A[9:0],
RAS_n/A CAS_n/A WE_n/ A10/ A[12:11], MR3[8:6
Refresh CS_n ACT_n 15 14 A13 BG1 BG0 AP A[20:16] ]
Fixed rate L H L L H V V V V 0vv
OTF: 1x L H L L H V L V V 1vv
OTF: 2x L H L L H V H V V 101
OTF: 4x L H L L H V H V V 110
tREFI and tRFC Parameters

The default refresh rate mode is fixed 1x mode where REFRESH commands should be
issued with the normal rate; that is, tREFI1 = tREFI(base) (for T C ื 85°C), and the dura-
tion of each REFRESH command is the normal REFRESH cycle time (tRFC1). In 2x
mode (either fixed 2x or OTF 2x mode), REFRESH commands should be issued to the
device at the double frequency (tREFI2 = tREFI(base)/2) of the normal refresh rate. In 4x
mode, the REFRESH command rate should be quadrupled (tREFI4 = tREFI(base)/4). Per
each mode and command type, the tRFC parameter has different values as defined in
the following table.
For discussion purposes, the REFRESH command that should be issued at the normal
refresh rate and has the normal REFRESH cycle duration may be referred to as an REF1x
command. The REFRESH command that should be issued at the double frequency
(tREFI2 = tREFI(base)/2) may be referred to as a REF2x command. Finally, the REFRESH
command that should be issued at the quadruple rate (tREFI4 = tREFI(base)/4) may be
referred to as a REF4x command.
In the fixed 1x refresh rate mode, only REF1x commands are permitted. In the fixed 2x
refresh rate mode, only REF2x commands are permitted. In the fixed 4x refresh rate
mode, only REF4x commands are permitted. When the on-the-fly 1x/2x refresh rate
mode is enabled, both REF1x and REF2x commands are permitted. When the OTF
1x/4x refresh rate mode is enabled, both REF1x and REF4x commands are permitted.
Table 50: tREFI and tRFC Parameters
Refresh Mode Parameter 2Gb 4Gb 8Gb Units

tREFI (base) 7.8 7.8 7.8 μs
1x mode tREFI1 -40°C ื TC ื 85°C tREFI(base) tREFI(base) tREFI(base) μs
85°C ื TC ื 95°C tREFI(base)/2 tREFI(base)/2 tREFI(base)/2 μs
tRFC1 160 260 350 ns
2x mode tREFI2 -40°C ื TC ื 85°C tREFI(base)/2 tREFI(base)/2 tREFI(base)/2 μs
tRFC2 110 160 260 ns
4x mode tREFI4 -40°C ื TC ื 85°C tREFI(base)/4 tREFI(base)/4 tREFI(base)/4 μs
tRFC4 90 110 160 ns
CCMTD-1725822587-10418
Figure 83: 4Gb with Fine Granularity Refresh Mode Example
Normal Temperature Operation – –40°C to 85°C Extended Temperature Operation – –40°C to 125°C
1x Mode 2x Mode 4x Mode 1x Mode 2x Mode 4x Mode

(–40°C to 85°C) (–40°C to 85°C) (–40°C to 85°C) (–40°C to 125°C) (–40°C to 125°C) (–40°C to 125°C)
s
5μ
REF@260ns REF@160ns REF@110ns REF@260ns REF@160ns REF@110ns
97
0.
μs
μs
=
95
95
FI
Rt E
1.
1.
s
5μ
=
=
REF@110ns
FI
FI
97
Rt E
Rt E
0.
s
=
9μ
9μ
FI
3.
3.
Rt E
s
=
5μ
REF@110ns REF@160ns REF@110ns
FI
FI
97
Rt E
Rt E
0.
μs
μs
=
95
95
FI
Rt E
1.
1.
s
5μ
=
=
REF@110ns
FI
FI
97
Rt E
Rt E
0.
=
FI
Rt E
s
8μ
s
5μ
REF@160ns REF@110ns REF@260ns REF@160ns REF@110ns
7.
97
=
0.
FI
μs
μs
=
Rt E
95
95
FI
Rt E
1.
1.
s
5μ
=
=
REF@110ns
FI
FI
97
Rt E
Rt E
0.
=
s
s
9μ
9μ
FI
3.
3.
Rt E
s
=
5μ
FI
FI
97
Rt E
Rt E
0.
μs
μs
=
FI
95
95
Rt E
1.
1.
s
5μ
=
=
REF@110ns
97
FI
FI
Rt E
Rt E
0.
=
FI
Rt E
s
5μ
148
97
0.
=
μs
μs
FI
95
95
Rt E
1.
1.
s
5μ
=
=
REF@110ns
97
FI
FI

Rt E
Rt E
0.
=
s
s
9μ
9μ
FI
Rt E
3.
3.
s
5μ
=
=
FI
FI
97
Rt E
Rt E
0.
=
μs
μs
FI
95
95
Rt E
1.
1.
s
5μ
s
8μ
=
REF@110ns
97
FI
FI
7.
0.
Rt E
Rt E
=
=
FI
FI
Rt E

Rt E
s
5μ
REF@160ns REF@110ns REF@260ns REF@160ns REF@110ns
97
0.
=
μs
μs
FI
95
95
Rt E
s
1.
1.
5μ
=
=
REF@110ns
97
FI
FI
0.
Rt E
Rt E
=
s
s
9μ
9μ
FI
Rt E
3.
3.
s
5μ
=
=
FI
FI
97
Rt E
Rt E
0.
=
μs
μs
FI
95
95
Rt E
s
1.
1.
5μ
REF@110ns
=
97
FI
FI
0.
Rt E
Rt E
=
FI
Rt E
Note: 1. tREFI value is dependent on operating temperature range. See Table 50.
Changing Refresh Rate

If the refresh rate is changed by either MRS or OTF. New tREFI and tRFC parameters will
be applied from the moment of the rate change. When the REF1x command is issued to
the DRAM, tREF1 and tRFC1 are applied from the time that the command was issued;
when the REF2x command is issued, tREF2 and tRFC2 should be satisfied.
Figure 84: OTF REFRESH Command Timing

CK_c
CK_t
Command DES REF1 DES DES DES Valid Valid REF2 DES DES Valid DES REF2 DES
tRFC1 (MIN) tRFC2 (MIN)
tREFI1 tREFI2
Don’t Care
The following conditions must be satisfied before the refresh rate can be changed. Oth-
erwise, data retention cannot be guaranteed.
• In the fixed 2x refresh rate mode or the OTF 1x/2x refresh mode, an even number of
REF2x commands must be issued because the last change of the refresh rate mode
with an MRS command before the refresh rate can be changed by another MRS com-
mand.
• In the OTF1x/2x refresh rate mode, an even number of REF2x commands must be is-
sued between any two REF1x commands.
• In the fixed 4x refresh rate mode or the OTF 1x/4x refresh mode, a multiple-of-four
number of REF4x commands must be issued because the last change of the refresh
rate with an MRS command before the refresh rate can be changed by another MRS
command.
• In the OTF1x/4x refresh rate mode, a multiple-of-four number of REF4x commands
must be issued between any two REF1x commands.
There are no special restrictions for the fixed 1x refresh rate mode. Switching between
fixed and OTF modes keeping the same rate is not regarded as a refresh rate change.
Usage with TCR Mode

If the temperature controlled refresh mode is enabled, only the normal mode (fixed 1x
mode, MR3[8:6] = 000) is allowed. If any other refresh mode than the normal mode is
selected, the temperature controlled refresh mode must be disabled.
Self Refresh Entry and Exit

The device can enter self refresh mode anytime in 1x, 2x, and 4x mode without any re-
striction on the number of REFRESH commands that have been issued during the
mode before the self refresh entry. However, upon self refresh exit, extra REFRESH com-
mand(s) may be required, depending on the condition of the self refresh entry.
The conditions and requirements for the extra REFRESH command(s) are defined as
follows:
• In the fixed 2x refresh rate mode or the enable-OTF 1x/2x refresh rate mode, it is rec-
ommended there be an even number of REF2x commands before entry into self re-
fresh after the last self refresh exit, REF1x command, or MRS command that set the
refresh mode. If this condition is met, no additional REFRESH commands are re-
quired upon self refresh exit. In the case that this condition is not met, either one ex-
tra REF1x command or two extra REF2x commands must be issued upon self refresh
exit. These extra REFRESH commands are not counted toward the computation of the
average refresh interval (tREFI).
• In the fixed 4x refresh rate mode or the enable-OTF 1x/4x refresh rate mode, it is rec-
ommended there be a multiple-of-four number of REF4x commands before entry in-
to self refresh after the last self refresh exit, REF1x command, or MRS command that
set the refresh mode. If this condition is met, no additional refresh commands are re-
quired upon self refresh exit. When this condition is not met, either one extra REF1x
command or four extra REF4x commands must be issued upon self refresh exit. These
extra REFRESH commands are not counted toward the computation of the average
refresh interval (tREFI).
There are no special restrictions on the fixed 1x refresh rate mode.
This section does not change the requirement regarding postponed REFRESH com-
mands. The requirement for the additional REFRESH command(s) described above is
independent of the requirement for the postponed REFRESH commands.
SELF REFRESH Operation

The SELF REFRESH command can be used to retain data in the device, even if the rest
of the system is powered down. When in self refresh mode, the device retains data with-
out external clocking. The device has a built-in timer to accommodate SELF REFRESH
operation. The SELF REFRESH command is defined by having CS_n, RAS_n, CAS_n,
and CKE held LOW with WE_n and ACT_n HIGH at the rising edge of the clock.
Before issuing the SELF REFRESH ENTRY command, the device must be idle with all
banks in the precharge state and tRP satisfied. Idle state is defined as: All banks are
closed (tRP, tDAL, and so on, satisfied), no data bursts are in progress, CKE is HIGH, and
all timings from previous operations are satisfied (tMRD, tMOD, tRFC, tZQinit, tZQoper,
tZQCS, and so on). After the SELF REFRESH ENTRY command is registered, CKE must
be held LOW to keep the device in self refresh mode. The DRAM automatically disables
ODT termination, regardless of the ODT pin, when it enters self refresh mode and auto-
matically enables ODT upon exiting self refresh. During normal operation (DLL_on),
the DLL is automatically disabled upon entering self refresh and is automatically ena-
bled (including a DLL reset) upon exiting self refresh.
When the device has entered self refresh mode, all of the external control signals, except
CKE and RESET_n, are “Don’t Care.” For proper SELF REFRESH operation, all power
supply and reference pins (VDD, V DDQ, V SS, V SSQ, V PP, and V REFCA) must be at valid levels.
The DRAM internal V REFDQ generator circuitry may remain on or be turned off depend-
ing on the MR6 bit 7 setting. If the internal V REFDQ circuit is on in self refresh, the first
WRITE operation or first write-leveling activity may occur after tXS time after self re-
fresh exit. If the DRAM internal V REFDQ circuitry is turned off in self refresh, it ensures
that the V REFDQ generator circuitry is powered up and stable within the tXSDLL period
when the DRAM exits the self refresh state. The first WRITE operation or first write-lev-
eling activity may not occur earlier than tXSDLL after exiting self refresh. The device ini-
tiates a minimum of one REFRESH command internally within the tCKE period once it
enters self refresh mode.
The clock is internally disabled during a SELF REFRESH operation to save power. The
minimum time that the device must remain in self refresh mode is tCKESR/
tCKESR_PAR. The user may change the external clock frequency or halt the external
clock tCKSRE/tCKSRE_PAR after self refresh entry is registered; however, the clock must
be restarted and tCKSRX must be stable before the device can exit SELF REFRESH oper-
ation.
The procedure for exiting self refresh requires a sequence of events. First, the clock must
be stable prior to CKE going back HIGH. Once a SELF REFRESH EXIT command (SRX,
combination of CKE going HIGH and DESELECT on the command bus) is registered,
the following timing delay must be satisfied:
Commands that do not require locked DLL:
• tXS = ACT, PRE, PREA, REF, SRE, and PDE.
• tXS_FAST = ZQCL, ZQCS, and MRS commands. For an MRS command, only DRAM
CL, WR/RTP register, and DLL reset in MR0; R TT(NOM) register in MR1; the CWL and
RTT(WR) registers in MR2; and gear-down mode register in MR3; WRITE and READ pre-
amble registers in MR4; RTT(PARK) register in MR5; tCCD_L/tDLLK and V REFDQ calibra-
tion value registers in MR6 may be accessed provided the DRAM is not in per-DRAM
mode. Access to other DRAM mode registers must satisfy tXS timing. WRITE com-
mands (WR, WRS4, WRS8, WRA, WRAS4, and WRAS8) that require synchronous ODT
and dynamic ODT controlled by the WRITE command require a locked DLL.
Commands that require locked DLL in the normal operating range:

• tXSDLL – RD, RDS4, RDS8, RDA, RDAS4, and RDAS8 (unlike DDR3, WR, WRS4, WRS8,
WRA, WRAS4, and WRAS8 because synchronous ODT is required).
Depending on the system environment and the amount of time spent in self refresh, ZQ
CALIBRATION commands may be required to compensate for the voltage and tempera-
ture drift described in the ZQ CALIBRATION Commands section. To issue ZQ CALIBRA-
TION commands, applicable timing requirements must be satisfied (see the ZQ Calibra-
tion Timing figure).
CKE must remain HIGH for the entire self refresh exit period tXSDLL for proper opera-
tion except for self refresh re-entry. Upon exit from self refresh, the device can be put
back into self refresh mode or power-down mode after waiting at least tXS period and
issuing one REFRESH command (refresh period of tRFC). The DESELECT command
must be registered on each positive clock edge during the self refresh exit interval tXS.
ODT must be turned off during tXSDLL.
The use of self refresh mode introduces the possibility that an internally timed refresh
event can be missed when CKE is raised for exit from self refresh mode. Upon exit from
self refresh, the device requires a minimum of one extra REFRESH command before it is
put back into self refresh mode.
Figure 85: Self Refresh Entry/Exit Timing

T0 T1 Ta0 Tb0 Tc0 Td0 Td1 Te0 Tf0 Tg0
CK_c
CK_t
tCKSRE/tCKSRE_PAR tCKSRX
tIS tCPDED
tCKESR/tCKESR_PAR
ODT Valid
tXS_FAST
Command DES SRE DES SRX Valid 1 Valid 2 Valid 3
ADDR Valid Valid Valid
tRP tXS
tXSDLL
Enter Self Refresh Exit Self Refresh

Notes: 1. Only MRS (limited to those described in the SELF REFRESH Operation section), ZQCS, or
ZQCL commands are allowed.
2. Valid commands not requiring a locked DLL.
Figure 86: Self Refresh Entry/Exit Timing with CAL Mode

7 7 7 7 7 7 7 7D 7D 7D 7D 7D 7E 7E 7E
&.BF
&.BW
W
&.65( W
&.65;
&6BQ
1RWH 1RWH
&RPPDQG '(6 '(6 65( '(6 '(6 65; '(6 '(6 '(6 '(6 9DOLG
ZR&6BQ
$''5 9DOLG 9DOLG

W
&$/ W
&3'(' W
;6B)$67 W
&$/
&.(
'RQ¶W&DUH
Notes: 1. tCAL = 3nCK, tCPDED = 4nCK, tCKSRE/tCKSRE_PAR = 8nCK, tCKSRX = 8nCK, tXS_FAST =
tREFC4 (MIN) + 10ns.
2. CS_n = HIGH, ACT_n = "Don't Care," RAS_n/A16 = "Don't Care," CAS_n/A15 = "Don't
Care," WE_n/A14 = "Don't Care."
3. Only MRS (limited to those described in the SELF REFRESH Operations section), ZQCS, or
4. The figure only displays tXS_FAST timing, but tCAL must also be added to any tXS and
tXSDLL associated commands during CAL mode.
Self Refresh Abort

The exit timing from self refresh exit to the first valid command not requiring a locked
DLL is tXS. The value of tXS is (tRFC+ 10ns). This delay allows any refreshes started by
the device time to complete. tRFC continues to grow with higher density devices, so tXS
will grow as well. An MRS bit enables the self refresh abort mode. If the bit is disabled,
the controller uses tXS timings (location MR4, bit 9). If the bit is enabled, the device
aborts any ongoing refresh and does not increment the refresh counter. The controller
can issue a valid command not requiring a locked DLL after a delay of tXS_ABORT.
Upon exit from self refresh, the device requires a minimum of one extra REFRESH com-
mand before it is put back into self refresh mode. This requirement remains the same
irrespective of the setting of the MRS bit for self refresh abort.
Figure 87: Self Refresh Abort

T0 T1 Ta0 Tb0 Tc0 Td0 Td1 Te0 Tf0 Tg0
CK_c
CK_t
tCKSRE/tCKSRE_PAR tCKSRX
tIS tCPDED

tCKESR/tCKESR_PAR
ODT Valid
tXS_FAST
Command DES SRE DES SRX Valid 1 Valid 2 Valid 3
ADDR Valid Valid Valid
tRP tXS_ABORT
tXSDLL
Enter Self Refresh Exit Self Refresh

Notes: 1. Only MRS (limited to those described in the SELF REFRESH Operation section), ZQCS, or
2. Valid commands not requiring a locked DLL with self refresh abort mode enabled in the
mode register.
Self Refresh Exit with NOP Command

Exiting self refresh mode using the NO OPERATION command (NOP) is allowed under a
specific system application. This special use of NOP allows for a common command/
address bus between active DRAM devices and DRAM(s) in maximum power saving
mode. Self refresh mode may exit with NOP commands provided:
• The device entered self refresh mode with CA parity, CAL, and gear-down disabled.
• tMPX_S and tMPX_LH are satisfied.
• NOP commands are only issued during tMPX_LH window.
No other command is allowed during the tMPX_LH window after an SELF REFRESH EX-
IT (SRX) command is issued.
Figure 88: Self Refresh Exit with NOP Command

7D 7D 7D 7D 7E 7E 7E 7E 7F 7F 7F 7F 7F 7G 7G 7G 7G 7H 7H
&.BF
&.BW
W
&.65;
&.(
2'7
9DOLG
W
03;B6 W
03;B/+
&6BQ
1RWH 1RWH
&RPPDQG 65; 123 123 123 123 '(6 '(6 '(6 '(6 '(6 9DOLG '(6 9DOLG
$''5 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG 9DOLG
W
;6
W
;6 W ;6'//
'RQ¶W&DUH
Power-Down Mode
Power-Down Mode
Power-down is synchronously entered when CKE is registered LOW (along with a DESE-
LECT command). CKE is not allowed to go LOW when the following operations are in
progress: MRS command, MPR operations, ZQCAL operations, DLL locking, or READ/
WRITE operations. CKE is allowed to go LOW while any other operations, such as ROW
ACTIVATION, PRECHARGE or auto precharge, or REFRESH, are in progress, but the
power-down IDD specification will not be applied until those operations are complete.
The timing diagrams that follow illustrate power-down entry and exit.
For the fastest power-down exit timing, the DLL should be in a locked state when pow-
er-down is entered. If the DLL is not locked during power-down entry, the DLL must be
reset after exiting power-down mode for proper READ operation and synchronous ODT
operation. DRAM design provides all AC and DC timing and voltage specification as
well as proper DLL operation with any CKE intensive operations as long as the control-
ler complies with DRAM specifications.
During power-down, if all banks are closed after any in-progress commands are com-
pleted, the device will be in precharge power-down mode; if any bank is open after in-
progress commands are completed, the device will be in active power-down mode.
Entering power-down deactivates the input and output buffers, excluding CK, CKE, and
RESET_n. In power-down mode, DRAM ODT input buffer deactivation is based on
Mode Register 5, bit 5 (MR5[5]). If it is configured to 0b, the ODT input buffer remains
on and the ODT input signal must be at valid logic level. If it is configured to 1b, the
ODT input buffer is deactivated and the DRAM ODT input signal may be floating and
the device does not provide RTT(NOM) termination. Note that the device continues to
provide RTT(Park) termination if it is enabled in MR5[8:6]. To protect internal delay on the
CKE line to block the input signals, multiple DES commands are needed during the CKE
switch off and on cycle(s); this timing period is defined as tCPDED. CKE LOW will result
in deactivation of command and address receivers after tCPDED has expired.
Table 51: Power-Down Entry Definitions
Power-
DRAM Status DLL Down Exit Relevant Parameters
Active On Fast tXP to any valid command.
(a bank or more open)
Precharged On Fast tXP to any valid command.
(all banks precharged)
The DLL is kept enabled during precharge power-down or active power-down. In pow-
er-down mode, CKE is LOW, RESET_n is HIGH, and a stable clock signal must be main-
tained at the inputs of the device. ODT should be in a valid state, but all other input sig-
nals are "Don't Care." (If RESET_n goes LOW during power-down, the device will be out
of power-down mode and in the reset state.) CKE LOW must be maintained until tCKE
has been satisfied. Power-down duration is limited by 9 × tREFI.
The power-down state is synchronously exited when CKE is registered HIGH (along
with DES command). CKE HIGH must be maintained until tCKE has been satisfied. The
ODT input signal must be at a valid level when the device exits from power-down mode,
independent of MR1 bit [10:8] if RTT(NOM) is enabled in the mode register. If RTT(NOM) is
disabled, the ODT input signal may remain floating. A valid, executable command can
Power-Down Mode
be applied with power-down exit latency, tXP, after CKE goes HIGH. Power-down exit la-
tency is defined in the AC Specifications table.
Figure 89: Active Power-Down Entry and Exit

T0 T1 T2 Ta0 Ta1 Tb0 Tb1 Tc0
CK_c
CK_t
Command Valid DES DES DES DES DES Valid
tPD
tIS
tIH
CKE Valid Valid
tCKE
tIH tIS
ODT (ODT buffer enabled - MR5[5] = 0)2
Refer to ODT Power-Down Entry/Exit

ODT (ODT buffer disabled - MR5[5] = 1)3 with ODT Buffer Disable Mode figures
Address Valid Valid
tCPDED tXP
Enter Exit
power-down power-down
mode mode
Notes: 1. Valid commands at T0 are ACT, DES, or PRE with one bank remaining open after comple-
tion of the PRECHARGE command.
2. ODT pin driven to a valid state; MR5[5] = 0 (normal setting).
3. ODT pin drive/float timing requirements for the ODT input buffer disable option (for ad-
ditional power savings during active power-down) is described in the section for ODT In-
put Buffer Disable Mode for Power-Down (page 164); MR5[5] = 1.
Power-Down Mode
Figure 90: Power-Down Entry After Read and Read with Auto Precharge
T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Tb0 Tb1
CK_c
CK_t
RD or DES DES DES DES DES DES DES DES DES Valid
Command RDA DES DES
tIS tCPDED
CKE Valid
Address Valid Valid
RL = AL + CL tPD
DQS_t, DQS_c
DI DI DI DI DI DI DI DI
DQ BL8 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
DQ BC4 DI DI DI DI
n n+1 n+2 n+3
tRDPDEN
Power-Down
entry
Transitioning Data Time Break Don’t Care
Note: 1. DI n (or b) = data-in from column n (or b).
Figure 91: Power-Down Entry After Write and Write with Auto Precharge
T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Tb0 Tb1 Tb2 Tc0 Tc1
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES Valid
tIS tCPDED
CKE Valid
Address Bank, Valid

Col n
A10
WL = AL + CWL WR tPD
DQS_t, DQS_c
DQ BL8 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
Start internal
precharge
DI DI DI DI
DQ BC4 n n+1 n+2 n+3
tWRAPDEN
Power-Down
entry
Transitioning Data 7LPH%UHDN 'RQ¶W&DUH
Notes: 1. DI n (or b) = data-in from column n (or b).

2. Valid commands at T0 are ACT, DES, or PRE with one bank remaining open after comple-
tion of the PRECHARGE command.
Power-Down Mode
Figure 92: Power-Down Entry After Write

T0 T1 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Tb0 Tb1 Tb2 Tc0 Tc1
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES Valid
tIS tCPDED
CKE Valid
Bank,
Address Col n Valid
A10
WL = AL + CWL tWR tPD
DQS_t, DQS_c
DQ BL8 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
DI DI DI DI
DQ BC4 n n+1 n+2 n+3
tWRPDEN
Power-Down
entry
Transitioning Data Time Break Don’t Care
Note: 1. DI n (or b) = data-in from column n (or b).
Figure 93: Precharge Power-Down Entry and Exit
T0 T1 T2 Ta0 Ta1 Tb0 Tb1 Tc0

CK_c
CK_t
Command DES DES DES DES DES DES Valid
tCPDED tCKE
tIS tIH
CKE Valid Valid
tPD tIS tXP
Enter Exit
power-down power-down
mode mode
Power-Down Mode
Figure 94: REFRESH Command to Power-Down Entry
T0 T1 T2 Ta0 Tb0 Tb1

CK_c
CK_t
Command REF DES DES DES DES
Address Valid
tCPDED
tIS tPD tCKE
CKE Valid
tREFPDEN
Figure 95: Active Command to Power-Down Entry

CK_c
CK_t
Command ACT DES DES DES DES
Address Valid
tCPDED
tIS tPD tCKE
CKE Valid
tACTPDEN
Power-Down Mode
Figure 96: PRECHARGE/PRECHARGE ALL Command to Power-Down Entry

CK_c
CK_t
Command PRE or
PREA DES DES DES Valid
Address Valid
tCPDED
tIS tPD tCKE
CKE
tPREPDEN
Figure 97: MRS Command to Power-Down Entry
T0 T1 Ta0 Ta1 Tb0 Tb1

CK_c
CK_t
Command MRS DES DES DES DES
Address Valid
tCPDED
tIS tPD tCKE
CKE Valid
tMRSPDEN
Power-Down Clarifications – Case 1

When CKE is registered LOW for power-down entry, tPD (MIN) must be satisfied before
CKE can be registered HIGH for power-down exit. The minimum value of parameter
tPD (MIN) is equal to the minimum value of parameter tCKE (MIN) as shown in the
Timing Parameters by Speed Bin table. A detailed example of Case 1 follows.
Power-Down Mode
Figure 98: Power-Down Entry/Exit Clarifications – Case 1
T0 T1 T2 Ta0 Ta1 Tb0 Tb1 Tb2

CK_c
CK_t
Command Valid DES DES DES DES DES DES

tPD tPD
tIS tIH tIS
CKE
tIH tIS tCKE
Address Valid
tCPDED tCPDED
Enter Exit Enter

power-down power-down power-down
mode mode mode
Power-Down Entry, Exit Timing with CAL

Command/Address latency is used and additional timing restrictions are required when
entering power-down, as noted in the following figures.
Figure 99: Active Power-Down Entry and Exit Timing with CAL
7 7 7D 7D 7D 7E 7E 7F 7F 7G 7G 7H
&.BF
&.BW
&6BQ
&RPPDQG '(6 '(6 9DOLG '(6 '(6 '(6 '(6 '(6 '(6 '(6 9DOLG
$GGUHVV 9DOLG 9DOLG
W &$/ W,+ W &3'(' W ;3 W &$/
W ,+ W ,6
W ,6 W 3'
&.(
Power-Down Mode
Figure 100: REFRESH Command to Power-Down Entry with CAL

7 7 7D 7E 7E 7F 7F 7G 7G 7H 7H 7I
&.BF
&.BW
&6BQ
&RPPDQG '(6 '(6 5() '(6 '(6 '(6 '(6 '(6 '(6 '(6 9DOLG
$GGUHVV 9DOLG 9DOLG
W &$/ W 5()3'(1 W &3'(' W ;3 W &$/
W ,+ W ,6
W ,6 W 3'
&.( W,+
ODT Input Buffer Disable Mode for Power-Down

DRAM does not provide RTT_NOM termination during power-down when ODT input
buffer deactivation mode is enabled in MR5 bit A5.
To account for DRAM internal delay on CKE line to disable the ODT buffer and block
the sampled output, the host controller must continuously drive ODT to either low or
high when entering power down (from tDODTLoff+1 prior to CKE low till tCPDED after
CKE low).
The ODT signal is allowed to float after tCPDEDmin has expired. In this mode, RTT_NOM
termination corresponding to sampled ODT at the input when CKE is registered low
(and tANPD before that) may be either RTT_NOM or RTT_PARK. tANPD is equal to (WL-1)
and is counted backwards from PDE.
Figure 101: ODT Power-Down Entry with ODT Buffer Disable Mode
diff_CK
CKE
tDODTLoff +1 tCPDED (MIN)
ODT Floating
tADC (MIN)
DRAM_RTT_sync
RTT(NOM) RTT(Park)
(DLL enabled)
CA parity disabled tCPDED
DODTLoff (MIN) + tADC (MAX)
DRAM_RTT_async
RTT(NOM) RTT(Park)
(DLL disabled)
tAONAS (MIN)
tCPDED (MIN) + tAOFAS (MAX)
Figure 102: ODT Power-Down Exit with ODT Buffer Disable Mode
diff_CK
CKE
ODT_A Floating
(DLL enabled)
tADC (MAX)
tXP
DRAM_RTT_A RTT(Park) RTT(NOM)
DODTLon
tADC (MIN)
ODT_B Floating
(DLL disabled)
tXP
DRAM_RTT_B RTT(Park) RTT(NOM)
tAONAS (MIN)
tAOFAS (MAX)
CRC Write Data Feature

CRC Write Data
The CRC write data feature takes the CRC generated data from the DRAM controller and
compares it to the internally CRC generated data and determines whether the two
match (no CRC error) or do not match (CRC error).
Figure 103: CRC Write Data Operation
DRAM Controller DRAM
Data Data
CRC CRC CRC

engine Data Code engine
CRC Code CRC Code
Compare
CRC
WRITE CRC DATA Operation

A DRAM controller generates a CRC checksum using a 72-bit CRC tree and forms the
write data frames, as shown in the following CRC data mapping tables for the x4, x8, and
x16 configurations. A x4 device has a CRC tree with 32 input data bits used, and the re-
maining upper 40 bits D[71:32] being 1s. A x8 device has a CRC tree with 64 input data
bits used, and the remaining upper 8 bits dependant upon whether DM_n/DBI_n is
used (1s are sent when not used). A x16 device has two identical CRC trees each, one for
the lower byte and one for the upper byte, with 64 input data bits used by each, and the
remaining upper 8 bits on each byte dependant upon whether DM_n/DBI_n is used (1s
are sent when not used). For a x8 and x16 DRAMs, the DRAM memory controller must
send 1s in transfer 9 location whether or not DM_n/DBI_n is used.
The DRAM checks for an error in a received code word D[71:0] by comparing the re-
ceived checksum against the computed checksum and reports errors using the
ALERT_n signal if there is a mismatch. The DRAM can write data to the DRAM core
without waiting for the CRC check for full writes when DM is disabled. If bad data is
written to the DRAM core, the DRAM memory controller will try to overwrite the bad
data with good data; this means the DRAM controller is responsible for data coherency
when DM is disabled. However, in the case where both CRC and DM are enabled via
MRS (that is, persistent mode), the DRAM will not write bad data to the core when a
CRC error is detected.
DBI_n and CRC Both Enabled

The DRAM computes the CRC for received written data D[71:0]. Data is not inverted
back based on DBI before it is used for computing CRC. The data is inverted back based
on DBI before it is written to the DRAM core.
DM_n and CRC Both Enabled

When both DM and write CRC are enabled in the DRAM mode register, the DRAM cal-
culates CRC before sending the write data into the array. If there is a CRC error, the
DRAM blocks the WRITE operation and discards the data. The Nonconsecutive WRITE
(BL8/BC4-OTF) with 2tCK Preamble and Write CRC in Same or Different Bank Group and
the WRITE (BL8/BC4-OTF/Fixed) with 1tCK Preamble and Write CRC in Same or Different
BankGroup figures in the WRITE Operation section show timing differences when DM
is enabled.
DM_n and DBI_n Conflict During Writes with CRC Enabled

Both write DBI_n and DM_n can not be enabled at the same time; read DBI_n and
DM_n can be enabled at the same time.
CRC and Write Preamble Restrictions

When write CRC is enabled:
• And 1tCK WRITE preamble mode is enabled, a tCCD_S or tCCD_L of 4 clocks is not
allowed.
• And 2tCK WRITE preamble mode is enabled, a tCCD_S or tCCD_L of 6 clocks is not
allowed.
CRC Simultaneous Operation Restrictions

When write CRC is enabled, neither MPR writes nor per-DRAM mode is allowed.
CRC Polynomial
The CRC polynomial used by DDR4 is the ATM-8 HEC, X8 + X2 + X1 + 1.
A combinatorial logic block implementation of this 8-bit CRC for 72 bits of data in-
cludes 272 two-input XOR gates contained in eight 6-XOR-gate-deep trees.
The CRC polynomial and combinatorial logic used by DDR4 is the same as used on
GDDR5.
The error coverage from the DDR4 polynomial used is shown in the following table.
Table 52: CRC Error Detection Coverage
Error Type Detection Capability

Random single-bit errors 100%
Random double-bit errors 100%
Table 52: CRC Error Detection Coverage (Continued)
Error Type Detection Capability

Random odd count errors 100%
Random multibit UI vertical column error 100%
detection excluding DBI bits
CRC Combinatorial Logic Equations
module CRC8_D72;
// polynomial: (0 1 2 8)
// data width: 72
// convention: the first serial data bit is D[71]
//initial condition all 0 implied
// "^" = XOR
function [7:0]
nextCRC8_D72;
input [71:0] Data;
input [71:0] D;
reg [7:0] CRC;
begin
D = Data;
CRC[0] =
D[69]^D[68]^D[67]^D[66]^D[64]^D[63]^D[60]^D[56]^D[54]^D[53]^D[52]^D[50]^D[49
]^D[48]^D[45]^D[43]^D[40]^D[39]^D[35]^D[34]^D[31]^D[30]^D[28]^D[23]^D[21]^D[1
9]^D[18]^D[16]^D[14]^D[12]^D[8]^D[7]^D[6]^D[0];
CRC[1] =
D[70]^D[66]^D[65]^D[63]^D[61]^D[60]^D[57]^D[56]^D[55]^D[52]^D[51]^D[48]^D[46
]^D[45]^D[44]^D[43]^D[41]^D[39]^D[36]^D[34]^D[32]^D[30]^D[29]^D[28]^D[24]^D[2
3]^D[22]^D[21]^D[20]^D[18]^D[17]^D[16]^D[15]^D[14]^D[13]^D[12]^D[9]^D[6]^D[1
]^D[0];
CRC[2] =
D[71]^D[69]^D[68]^D[63]^D[62]^D[61]^D[60]^D[58]^D[57]^D[54]^D[50]^D[48]^D[47
]^D[46]^D[44]^D[43]^D[42]^D[39]^D[37]^D[34]^D[33]^D[29]^D[28]^D[25]^D[24]^D[2
2]^D[17]^D[15]^D[13]^D[12]^D[10]^D[8]^D[6]^D[2]^D[1]^D[0];
CRC[3] =
D[70]^D[69]^D[64]^D[63]^D[62]^D[61]^D[59]^D[58]^D[55]^D[51]^D[49]^D[48]^D[47
]^D[45]^D[44]^D[43]^D[40]^D[38]^D[35]^D[34]^D[30]^D[29]^D[26]^D[25]^D[23]^D[1
8]^D[16]^D[14]^D[13]^D[11]^D[9]^D[7]^D[3]^D[2]^D[1];
CRC[4] =
D[71]^D[70]^D[65]^D[64]^D[63]^D[62]^D[60]^D[59]^D[56]^D[52]^D[50]^D[49]^D[48
]^D[46]^D[45]^D[44]^D[41]^D[39]^D[36]^D[35]^D[31]^D[30]^D[27]^D[26]^D[24]^D[1
9]^D[17]^D[15]^D[14]^D[12]^D[10]^D[8]^D[4]^D[3]^D[2];
CRC[5] =
D[71]^D[66]^D[65]^D[64]^D[63]^D[61]^D[60]^D[57]^D[53]^D[51]^D[50]^D[49]^D[47
]^D[46]^D[45]^D[42]^D[40]^D[37]^D[36]^D[32]^D[31]^D[28]^D[27]^D[25]^D[20]^D[1
8]^D[16]^D[15]^D[13]^D[11]^D[9]^D[5]^D[4]^D[3];
CRC[6] =
D[67]^D[66]^D[65]^D[64]^D[62]^D[61]^D[58]^D[54]^D[52]^D[51]^D[50]^D[48]^D[47
]^D[46]^D[43]^D[41]^D[38]^D[37]^D[33]^D[32]^D[29]^D[28]^D[26]^D[21]^D[19]^D[1
7]^D[16]^D[14]^D[12]^D[10]^D[6]^D[5]^D[4];
CRC[7] =
D[68]^D[67]^D[66]^D[65]^D[63]^D[62]^D[59]^D[55]^D[53]^D[52]^D[51]^D[49]^D[48
]^D[47]^D[44]^D[42]^D[39]^D[38]^D[34]^D[33]^D[30]^D[29]^D[27]^D[22]^D[20]^D[1
8]^D[17]^D[15]^D[13]^D[11]^D[7]^D[6]^D[5];
nextCRC8_D72 = CRC;
Burst Ordering for BL8

DDR4 supports fixed WRITE burst ordering [A2:A1:A0 = 0:0:0] when write CRC is ena-
bled in BL8 (fixed).
CRC Data Bit Mapping
Table 53: CRC Data Mapping for x4 Devices, BL8
Func- Transfer
tion 0 1 2 3 4 5 6 7 8 9
DQ0 D0 D1 D2 D3 D4 D5 D6 D7 CRC0 CRC4
Func- Transfer
tion 0 1 2 3 4 5 6 7 8 9
DQ0 D0 D1 D2 D3 D4 D5 D6 D7 CRC0 1
DQ1 D8 D9 D10 D11 D12 D13 D14 D15 CRC1 1
DQ2 D16 D17 D18 D19 D20 D21 D22 D23 CRC2 1
DQ3 D24 D25 D26 D27 D28 D29 D30 D31 CRC3 1
DQ4 D32 D33 D34 D35 D36 D37 D38 D39 CRC4 1
DQ5 D40 D41 D42 D43 D44 D45 D46 D47 CRC5 1
DQ6 D48 D49 D50 D51 D52 D53 D54 D55 CRC6 1
DQ7 D56 D57 D58 D59 D60 D61 D62 D63 CRC7 1
DM_n/ D64 D65 D66 D67 D68 D69 D70 D71 1 1
DBI_n
A x16 device is treated as two x8 devices; a x16 device will have two identical CRC trees
implemented. CRC[7:0] covers data bits D[71:0], and CRC[15:8] covers data bits
D[143:72].
Func- Transfer
tion 0 1 2 3 4 5 6 7 8 9
DQ0 D0 D1 D2 D3 D4 D5 D6 D7 CRC0 1
DQ1 D8 D9 D10 D11 D12 D13 D14 D15 CRC1 1
DQ2 D16 D17 D18 D19 D20 D21 D22 D23 CRC2 1
DQ3 D24 D25 D26 D27 D28 D29 D30 D31 CRC3 1
DQ4 D32 D33 D34 D35 D36 D37 D38 D39 CRC4 1
DQ5 D40 D41 D42 D43 D44 D45 D46 D47 CRC5 1
DQ6 D48 D49 D50 D51 D52 D53 D54 D55 CRC6 1
DQ7 D56 D57 D58 D59 D60 D61 D62 D63 CRC7 1
LDM_n/ D64 D65 D66 D67 D68 D69 D70 D71 1 1
LDBI_n
DQ8 D72 D73 D74 D75 D76 D77 D78 D79 CRC8 1
DQ9 D80 D81 D82 D83 D84 D85 D86 D87 CRC9 1
DQ10 D88 D89 D90 D91 D92 D93 D94 D95 CRC10 1
DQ11 D96 D97 D98 D99 D100 D101 D102 D103 CRC11 1
DQ12 D104 D105 D106 D107 D108 D109 D110 D111 CRC12 1
DQ13 D112 D113 D114 D115 D116 D117 D118 D119 CRC13 1
DQ14 D120 D121 D122 D123 D124 D125 D126 D127 CRC14 1
DQ15 D128 D129 D130 D131 D132 D133 D134 D135 CRC15 1
UDM_n/ D136 D137 D138 D139 D140 D141 D142 D143 1 1
UDBI_n
CRC Enabled With BC4

If CRC and BC4 are both enabled, then address bit A2 is used to transfer critical data
first for BC4 writes.
CRC with BC4 Data Bit Mapping

For a x4 device, the CRC tree inputs are 16 data bits, and the inputs for the remaining
bits are 1.
When A2 = 1, data bits D[7:4] are used as inputs for D[3:0], D[15:12] are used as inputs to
D[11:8], and so forth, for the CRC tree.
Table 56: CRC Data Mapping for x4 Devices, BC4
Transfer
Function 0 1 2 3 4 5 6 7 8 9
A2 = 0
DQ0 D0 D1 D2 D3 1 1 1 1 CRC0 CRC4
DQ1 D8 D9 D10 D11 1 1 1 1 CRC1 CRC5
DQ2 D16 D17 D18 D19 1 1 1 1 CRC2 CRC6
Table 56: CRC Data Mapping for x4 Devices, BC4 (Continued)
Transfer
Function 0 1 2 3 4 5 6 7 8 9
DQ3 D24 D25 D26 D27 1 1 1 1 CRC3 CRC7
A2 = 1
DQ0 D4 D5 D6 D7 1 1 1 1 CRC0 CRC4
DQ1 D12 D13 D14 D15 1 1 1 1 CRC1 CRC5
DQ2 D20 D21 D22 D23 1 1 1 1 CRC2 CRC6
DQ3 D28 D29 D30 D31 1 1 1 1 CRC3 CRC7
For a x8 device, the CRC tree inputs are 36 data bits.

When A2 = 0, the input bits D[67:64]) are used if DBI_n or DM_n functions are enabled;
if DBI_n and DM_n are disabled, then D[67:64]) are 1.
When A2 = 1, data bits D[7:4] are used as inputs for D[3:0], D[15:12] are used as inputs to
D[11:8], and so forth, for the CRC tree. The input bits D[71:68]) are used if DBI_n or
DM_n functions are enabled; if DBI_n and DM_n are disabled, then D[71:68]) are 1.
Transfer
Function 0 1 2 3 4 5 6 7 8 9
A2 = 0
DQ0 D0 D1 D2 D3 1 1 1 1 CRC0 1
DQ1 D8 D9 D10 D11 1 1 1 1 CRC1 1
DQ2 D16 D17 D18 D19 1 1 1 1 CRC2 1
DQ3 D24 D25 D26 D27 1 1 1 1 CRC3 1
DQ4 D32 D33 D34 D35 1 1 1 1 CRC4 1
DQ5 D40 D41 D42 D43 1 1 1 1 CRC5 1
DQ6 D48 D49 D50 D51 1 1 1 1 CRC6 1
DQ7 D56 D57 D58 D59 1 1 1 1 CRC7 1
DM_n/DBI_n D64 D65 D66 D67 1 1 1 1 1 1
A2 = 1
DQ0 D4 D5 D6 D7 1 1 1 1 CRC0 1
DQ1 D12 D13 D14 D15 1 1 1 1 CRC1 1
DQ2 D20 D21 D22 D23 1 1 1 1 CRC2 1
DQ3 D28 D29 D30 D31 1 1 1 1 CRC3 1
DQ4 D36 D37 D38 D39 1 1 1 1 CRC4 1
DQ5 D44 D45 D46 D47 1 1 1 1 CRC5 1
DQ6 D52 D53 D54 D55 1 1 1 1 CRC6 1
DQ7 D60 D61 D62 D63 1 1 1 1 CRC7 1
DM_n/DBI_n D68 D69 D70 D71 1 1 1 1 1 1
There are two identical CRC trees for x16 devices, each have CRC tree inputs of 36 bits.
When A2 = 0, input bits D[67:64] are used if DBI_n or DM_n functions are enabled; if
DBI_n and DM_n are disabled, then D[67:64] are 1s. The input bits D[139:136] are used
if DBI_n or DM_n functions are enabled; if DBI_n and DM_n are disabled, then
D[139:136] are 1s.
When A2 = 1, data bits D[7:4] are used as inputs for D[3:0], D[15:12] are used as inputs
for D[11:8], and so forth, for the CRC tree. Input bits D[71:68] are used if DBI_n or DM_n
functions are enabled; if DBI_n and DM_n are disabled, then D[71:68] are 1s. The input
bits D[143:140] are used if DBI_n or DM_n functions are enabled; if DBI_n and DM_n
are disabled, then D[143:140] are 1s.
Transfer
Function 0 1 2 3 4 5 6 7 8 9
A2 = 0
DQ0 D0 D1 D2 D3 1 1 1 1 CRC0 1
DQ1 D8 D9 D10 D11 1 1 1 1 CRC1 1
DQ2 D16 D17 D18 D19 1 1 1 1 CRC2 1
DQ3 D24 D25 D26 D27 1 1 1 1 CRC3 1
DQ4 D32 D33 D34 D35 1 1 1 1 CRC4 1
DQ5 D40 D41 D42 D43 1 1 1 1 CRC5 1
DQ6 D48 D49 D50 D51 1 1 1 1 CRC6 1
DQ7 D56 D57 D58 D59 1 1 1 1 CRC7 1
LDM_n/LDBI_n D64 D65 D66 D67 1 1 1 1 1 1
DQ8 D72 D73 D74 D75 1 1 1 1 CRC8 1
DQ9 D80 D81 D82 D83 1 1 1 1 CRC9 1
DQ10 D88 D89 D90 D91 1 1 1 1 CRC10 1
DQ11 D96 D97 D98 D99 1 1 1 1 CRC11 1
DQ12 D104 D105 D106 D107 1 1 1 1 CRC12 1
DQ13 D112 D113 D114 D115 1 1 1 1 CRC13 1
DQ14 D120 D121 D122 D123 1 1 1 1 CRC14 1
DQ15 D128 D129 D130 D131 1 1 1 1 CRC15 1
UDM_n/UDBI_n D136 D137 D138 D139 1 1 1 1 1 1
A2 = 1
DQ0 D4 D5 D6 D7 1 1 1 1 CRC0 1
DQ1 D12 D13 D14 D15 1 1 1 1 CRC1 1
DQ2 D20 D21 D22 D23 1 1 1 1 CRC2 1
DQ3 D28 D29 D30 D31 1 1 1 1 CRC3 1
DQ4 D36 D37 D38 D39 1 1 1 1 CRC4 1
DQ5 D44 D45 D46 D47 1 1 1 1 CRC5 1
DQ6 D52 D53 D54 D55 1 1 1 1 CRC6 1
DQ7 D60 D61 D62 D63 1 1 1 1 CRC7 1
LDM_n/LDBI_n D68 D69 D70 D71 1 1 1 1 1 1
Table 58: CRC Data Mapping for x16 Devices, BC4 (Continued)
Transfer
Function 0 1 2 3 4 5 6 7 8 9
DQ8 D76 D77 D78 D79 1 1 1 1 CRC8 1
DQ9 D84 D85 D86 D87 1 1 1 1 CRC9 1
DQ10 D92 D93 D94 D95 1 1 1 1 CRC10 1
DQ11 D100 D101 D102 D103 1 1 1 1 CRC11 1
DQ12 D108 D109 D110 D111 1 1 1 1 CRC12 1
DQ13 D116 D117 D118 D119 1 1 1 1 CRC13 1
DQ14 D124 D125 D126 D127 1 1 1 1 CRC14 1
DQ15 D132 D133 D134 D135 1 1 1 1 CRC15 1
UDM_n/UDBI_n D140 D141 D142 D143 1 1 1 1 1 1
CRC Equations for x8 Device in BC4 Mode with A2 = 0 and A2 = 1

The following example is of a CRC tree when x8 is used in BC4 mode (x4 and x16 CRC
trees have similar differences).
CRC[0], A2=0 =
1^1^D[67]^D[66]^D[64]^1^1^D[56]^1^1^1^D[50]^D[49]^D[48]^1^D[43]^D[40]^1^D[3
5]^D[34]^1^1^1^1^1^D[19]^D[18]^D[16]^1^1^D[8] ^1^1^ D[0] ;
CRC[0], A2=1 =
1^1^D[71]^D[70]^D[68]^1^1^D[60]^1^1^1^D[54]^D[53]^D[52]^1^D[47]^D[44]^1^D[3
9]^D[38]^1^1^1^1^1^D[23]^D[22]^D[20]^1^1^D[12]^1^1^D[4] ;
CRC[1], A2=0 =
1^D[66]^D[65]^1^1^1^D[57]^D[56]^1^1^D[51]^D[48]^1^1^1^D[43]^D[41]^1^1^D[34
]^D[32]^1^1^1^D[24]^1^1^1^1^D[18]^D[17]^D[16]^1^1^1^1^D[9] ^1^ D[1]^D[0];
CRC[1], A2=1 =
1^D[70]^D[69]^1^1^1^D[61]^D[60]^1^1^D[55]^D[52]^1^1^1^D[47]^D[45]^1^1^D[38
]^D[36]^1^1^1^D[28]^1^1^1^1^D[22]^D[21]^D[20]^1^1^1^1^D[13]^1^D[5]^D[4];
CRC[2], A2=0 =
1^1^1^1^1^1^1^D[58]^D[57]^1^D[50]^D[48]^1^1^1^D[43]^D[42]^1^1^D[34]^D[33]^1
^1^D[25]^D[24]^1^D[17]^1^1^1^D[10]^D[8] ^1^D[2]^D[1]^D[0];
CRC[2], A2=1 =
1^1^1^1^1^1^1^D[62]^D[61]^1^D[54]^D[52]^1^1^1^D[47]^D[46]^1^1^D[38]^D[37]^1
^1^D[29]^D[28]^1^D[21]^1^1^1^D[14]^D12]^1^D[6]^D[5]^D[4];
CRC[3], A2=0 =
1^1^D[64]^1^1^1^D[59]^D[58]^1^D[51]^D[49]^D[48]^1^1^1^D[43]^D[40]^1^D[35]^
D[34]^1^1^D[26]^D[25]^1^D[18]^D[16]^1^1^D[11]^D[9] ^1^D[3]^D[2]^D[1];
CRC[3], A2=1 =
1^1^D[68]^1^1^1^D[63]^D[62]^1^D[55]^D[53]^D[52]^1^1^1^D[47]^D[44]^1^D[39]^
D[38]^1^1^D[30]^D[29]^1^D[22]^D[20]^1^1^D[15]^D[13]^1^D[7]^D[6]^D[5];
CRC[4], A2=0 =
1^1^D[65]^D[64]^1^1^1^D[59]^D[56]^1^D[50]^D[49]^D[48]^1^1^1^D[41]^1^1^D[35
]^1^1^D[27]^D[26]^D[24]^D[19]^D[17]^1^1^1^D[10]^D[8] ^1^D[3]^D[2];
CRC[4], A2=1 =
1^1^D[69]^D[68]^1^1^1^D[63]^D[60]^1^D[54]^D[53]^D[52]^1^1^1^D[45]^1^1^D[39
]^1^1^D[31]^D[30]^D[28]^D[23]^D[21]^1^1^1^D[14]^D[12]^1^D[7]^D[6];
CRC[5], A2=0 =
1^D[66]^D[65]^D[64]^1^1^1^D[57]^1^D[51]^D[50]^D[49]^1^1^1^D[42]^D[40]^1^1^
D[32]^1^1^D[27]^D[25]^1^D[18]^D[16]^1^1^D[11]^D[9] ^1^1^D[3];
CRC[5], A2=1 =
1^D[70]^D[69]^D[68]^1^1^1^D[61]^1^D[55]^D[54]^D[53]^1^1^1^D[46]^D[44]^1^1^
D[36]^1^1^D[31]^D[29]^1^D[22]^D[20]^1^1^D[15]^D[13]^1^1^D[7];
CRC[6], A2=0 =
D[67]^D[66]^D[65]^D[64]^1^1^D[58]^1^1^D[51]^D[50]^D[48]^1^1^D[43]^D[41]^1^1
^D[33]^D[32]^1^1^D[26]^1^D[19]^D[17]^D[16]^1^1^D[10]^1^1^1;
CRC[6], A2=1 =
D[71]^D[70]^D[69]^D[68]^1^1^D[62]^1^1^D[55]^D[54]^D[52]^1^1^D[47]^D[45]^1^1
^D[37]^D[36]^1^1^D[30]^1^D[23]^D[21]^D[20]^1^1^D[14]^1^1^1;
CRC[7], A2=0 =
1^D[67]^D[66]^D[65]^1^1^D[59]^1^1^1^D[51]^D[49]^D[48]^1^1^D[42]^1^1^D[34]^
D[33]^1^1^D[27]^1^1^D[18]^D[17]^1^1^D[11]^1^1^1;
CRC[7], A2=1 =
1^D[71]^D[70]^D[69]^1^1^D[63]^1^1^1^D[55]^D[53]^D[52]^1^1^D[46]^1^1^D[38]^
D[37]^1^1^D[31]^1^1^D[22]^D[21]^1^1^D[15]^1^1^1;
CRC Error Handling

The CRC error mechanism shares the same ALERT_n signal as CA parity for reporting
write errors to the DRAM. The controller has two ways to distinguish between CRC er-
rors and CA parity errors: 1) Read DRAM mode/MPR registers, and 2) Measure time
ALERT_n is LOW. To speed up recovery for CRC errors, CRC errors are only sent back as
a "short" pulse; the maximum pulse width is roughly ten clocks (unlike CA parity where
ALERT_n is LOW longer than 45 clocks). The ALERT_n LOW could be longer than the
maximum limit at the controller if there are multiple CRC errors as the ALERT_n signals
are connected by a daisy chain bus. The latency to ALERT_n signal is defined as
tCRC_ALERT in the following figure.
The DRAM will set the error status bit located at MR5[3] to a 1 upon detecting a CRC
error, which will subsequently set the CRC error status flag in the MPR error log HIGH
(MPR Page1, MPR3[7]). The CRC error status bit (and CRC error status flag) remains set
at 1 until the DRAM controller clears the CRC error status bit using an MRS command
to set MR5[3] to a 0. The DRAM controller, upon seeing an error as a pulse width, will
retry the write transactions. The controller should consider the worst-case delay for
ALERT_n (during initialization) and backup the transactions accordingly. The DRAM
controller may also be made more intelligent and correlate the write CRC error to a spe-
cific rank or a transaction.
Figure 104: CRC Error Reporting

T0 T1 T2 T3 T4 T5 T6 Ta0 Ta1 Ta2 Ta3 Tb0 Tb1
CK_c
CK_t
DQIN
Dx Dx+1 Dx+2 Dx+3 Dx+4 Dx+5 Dx+6 Dx+7 CRCy 1
CRC ALERT_PW (MAX)

tCRC_ALERT
ALERT_n
CRC ALERT_PW (MIN)
Transition Data Don’t Care
Notes: 1. D[71:1] CRC computed by DRAM did not match CRC[7:0] at T5 and started error generat-
ing process at T6.
2. CRC ALERT_PW is specified from the point where the DRAM starts to drive the signal
LOW to the point where the DRAM driver releases and the controller starts to pull the
signal up.
3. Timing diagram applies to x4, x8, and x16 devices.
CCMTD-1725822587-10418
CRC Write Data Flow Diagram
Figure 105: CA Parity Flow Diagram
DRAM write MR2 12 enable CRC

process start MR5 3 set CRC error clear to 0
MR5 10 enable/disable DM
MR3[10:9] WCL if DM enabled
Capture data
Persistent DRAM
CRC Yes Yes CRC same as No
mode
enabled controller
enabled
CRC
No Yes
No
Transfer Data Transfer data Transfer data

Internally internally internally
Yes DRAM
CRC same as No MR5[3] = 0 Yes MR5[3] = 0 Yes
CA error controller at WRITE at WRITE
CRC
176
Yes ALERT_n LOW No Set error flag ALERT_n LOW No Set error flag
No 6 to 10 CKs MR5[A3] 1 6 to 10 CKs MR5[A3] 1
MR5[A3] and MR5[A3] and

PAGE1 MPR3[7] PAGE1 MPR3[7]
remain set to 1 Set error status remain set to 1 Set error status
ALERT_n HIGH ALERT_n HIGH
PAGE1 MPR3[7] 1 PAGE1 MPR3[7] 1
WRITE burst WRITE burst WRITE burst WRITE burst WRITE burst WRITE burst
completed completed completed completed completed rejected
Bad data written Bad data not written
MR5 3 reset to 0 if desired MR5 3 reset to 0 if desired

Data Bus Inversion
Data Bus Inversion

The DATA BUS INVERSION (DBI) function is supported only for x8 and x16 configura-
tions (it is not supported on x4 devices). DBI opportunistically inverts data bits, and in
conjunction with the DBI_n I/O, less than half of the DQs will switch LOW for a given
DQS strobe edge. The DBI function shares a common pin with the DATA MASK (DM)
and TDQS functions. The DBI function applies to either or both READ and WRITE oper-
ations: Write DBI cannot be enabled at the same time the DM function is enabled, and
DBI is not allowed during MPR READ operation. Valid configurations for TDQS, DM,
and DBI functions are shown below.
Table 59: DBI vs. DM vs. TDQS Function Matrix
Read DBI Write DBI Data Mask (DM) TDQS (x8 only)
Enabled (or Disabled) Disabled Disabled Disabled
MR5[12]=1 (or MR5[11] = 0 MR5[10] = 0 MR1[11] = 0
MR5[12] = 0) Enabled Disabled Disabled
MR5[11] = 1 MR5[10] = 0 MR1[11] = 0
Disabled Enabled Disabled
MR5[11] = 0 MR5[10] = 1 MR1[11] = 0
Disabled Disabled Disabled Enabled
MR5[12] = 0 MR5[11] = 0 MR5[10] = 0 MR1[11] = 1
DBI During a WRITE Operation

If DBI_n is sampled LOW on a given byte lane during a WRITE operation, the DRAM in-
verts write data received on the DQ inputs prior to writing the internal memory array. If
DBI_n is sampled HIGH on a given byte lane, the DRAM leaves the data received on the
DQ inputs noninverted. The write DQ frame format is shown below for x8 and x16 con-
figurations (the x4 configuration does not support the DBI function).
Table 60: DBI Write, DQ Frame Format (x8)
Transfer
Function 0 1 2 3 4 5 6 7
DQ[7:0] Byte 0 Byte 1 Byte 2 Byte 3 Byte 4 Byte 5 Byte 6 Byte 7
DM_n or DM0 or DM1 or DM2 or DM3 or DM4 or DM5 or DM6 or DM7 or
DBI_n DBI0 DBI1 DBI2 DBI3 DBI4 DBI5 DBI6 DBI7
Table 61: DBI Write, DQ Frame Format (x16)
Transfer, Lower (L) and Upper(U)

Function 0 1 2 3 4 5 6 7
DQ[7:0] LByte 0 LByte 1 LByte 2 LByte 3 LByte 4 LByte 5 LByte 6 LByte 7
LDM_n or LDM0 or LDM1 or LDM2 or LDM3 or LDM4 or LDM5 or LDM6 or LDM7 or
LDBI_n LDBI0 LDBI1 LDBI2 LDBI3 LDBI4 LDBI5 LDBI6 LDBI7
DQ[15:8] UByte 0 UByte 1 UByte 2 UByte 3 UByte 4 UByte 5 UByte 6 UByte 7
Data Bus Inversion
Table 61: DBI Write, DQ Frame Format (x16) (Continued)
Transfer, Lower (L) and Upper(U)

Function 0 1 2 3 4 5 6 7
UDM_n or UDM0 or UDM1 or UDM2 or UDM3 or UDM4 or UDM5 or UDM6 or UDM7 or
UDBI_n UDBI0 UDBI1 UDBI2 UDBI3 UDBI4 UDBI5 UDBI6 UDBI7
DBI During a READ Operation

If the number of 0 data bits within a given byte lane is greater than four during a READ
operation, the DRAM inverts read data on its DQ outputs and drives the DBI_n pin
LOW; otherwise, the DRAM does not invert the read data and drives the DBI_n pin
HIGH. The read DQ frame format is shown below for x8 and x16 configurations (the x4
configuration does not support the DBI function).
Table 62: DBI Read, DQ Frame Format (x8)
Transfer Byte
Function 0 1 2 3 4 5 6 7
Table 63: DBI Read, DQ Frame Format (x16)
Transfer Byte, Lower (L) and Upper(U)

Function 0 1 2 3 4 5 6 7
Data Mask
Data Mask
The DATA MASK (DM) function, also described as PARTIAL WRITE, is supported only
for x8 and x16 configurations (it is not supported on x4 devices). The DM function
shares a common pin with the DBI_n and TDQS functions. The DM function applies
only to WRITE operations and cannot be enabled at the same time the WRITE DBI
function is enabled. The valid configurations for the TDQS, DM, and DBI functions are
shown here.
Table 64: DM vs. TDQS vs. DBI Function Matrix
Data Mask (DM) TDQS (x8 only) Write DBI Read DBI
Enabled Disabled Disabled Enabled or Disabled
MR5[10] = 1 MR1[11] = 0 MR5[11] = 0 MR5[12] = 1 or
MR5[12] = 0
Disabled Enabled Disabled Disabled
MR5[10] = 0 MR1[11] = 1 MR5[11] = 0 MR5[12] = 0
Disabled Enabled Enabled or Disabled
MR1[11] = 0 MR5[11] = 1 MR5[12] = 1 or
MR5[12] = 0
Disabled Disabled Enabled (or Disabled)
MR1[11] = 0 MR5[11] = 0 MR5[12] = 1 (or
MR5[12] = 0)
When enabled, the DM function applies during a WRITE operation. If DM_n is sampled
LOW on a given byte lane, the DRAM masks the write data received on the DQ inputs. If
DM_n is sampled HIGH on a given byte lane, the DRAM does not mask the data and
writes this data into the DRAM core. The DQ frame format for x8 and x16 configurations
is shown below. If both CRC write and DM are enabled (via MRS), the CRC will be
checked and valid prior to the DRAM writing data into the DRAM core. If a CRC error
occurs while the DM feature is enabled, CRC write persistent mode will be enabled and
data will not be written into the DRAM core. In the case of CRC write enabled and DM
disabled (via MRS), that is, CRC write nonpersistent mode, data is written to the DRAM
core even if a CRC error occurs.
Table 65: Data Mask, DQ Frame Format (x8)
Transfer
Function 0 1 2 3 4 5 6 7
DM_n or DM0 or DM1 or DM2 or DM3 or DM4 or DM5 or DM6 or DM7 or
Table 66: Data Mask, DQ Frame Format (x16)
Transfer, Lower (L) and Upper (U)

Function 0 1 2 3 4 5 6 7
Data Mask
Table 66: Data Mask, DQ Frame Format (x16) (Continued)
Transfer, Lower (L) and Upper (U)

Function 0 1 2 3 4 5 6 7
LDM_n or LDM0 or LDM1 or LDM2 or LDM3 or LDM4 or LDM5 or LDM6 or LDM7 or
UDM_n or UDM0 or UDM1 or UDM2 or UDM3 or UDM4 or UDM5 or UDM6 or UDM7 or
Programmable Preamble Modes and DQS Postambles

The device supports programmable WRITE and READ preamble modes, either the nor-
mal 1tCK preamble mode or special 2tCK preamble mode. The 2 tCK preamble mode
places special timing constraints on many operational features as well as being suppor-
ted for data rates of DDR4-2400 and faster. The WRITE preamble 1 tCK or 2tCK mode
can be selected independently from READ preamble 1tCK or 2tCK mode.
READ preamble training is also supported; this mode can be used by the DRAM con-
troller to train or "read level" the DQS receivers.
There are tCCD restrictions under some circumstances:
• When 2tCK READ preamble mode is enabled, a tCCD_S or tCCD_L of 5 clocks is not
allowed.
• When 2tCK WRITE preamble mode is enabled and write CRC is not enabled, a tCCD_S
or tCCD_L of 5 clocks is not allowed.
• When 2tCK WRITE preamble mode is enabled and write CRC is enabled, a tCCD_S or
tCCD_L of 6 clocks is not allowed.
WRITE Preamble Mode

MR4[12] = 0 selects 1tCK WRITE preamble mode while MR4[12] = 1 selects 2 tCK WRITE
preamble mode. Examples are shown in the figures below.
Figure 106: 1tCK vs. 2tCK WRITE Preamble Mode
1tCK Mode
WR
WL
CK_c
CK_t
Preamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
2tCK Mode
WR
WL
CK_c
CK_t
Preamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
CWL has special considerations when in the 2tCK WRITE preamble mode. The CWL val-
ue selected in MR2[5:3], as seen in table below, requires at least one additional clock
when the primary CWL value and 2tCK WRITE preamble mode are used; no additional
clocks are required when the alternate CWL value and 2tCK WRITE preamble mode are
used.
Table 67: CWL Selection
CWL - Primary Choice CWL - Alternate Choice

Speed Bin 1tCK Preamble 2tCK Preamble 1tCK Preamble 2tCK Preamble
DDR4-1600 9 N/A 11 N/A
DDR4-1866 10 N/A 12 N/A
DDR4-2133 11 N/A 14 N/A
DDR4-2400 12 14 16 16
DDR4-2666 14 16 18 18
DDR4-2933 16 18 20 20
DDR4-3200 16 18 20 20
Note: 1. CWL programmable requirement for MR2[5:3].
When operating in 2tCK WRITE preamble mode, tWTR (command based) and tWR
(MR0[11:9]) must be programmed to a value 1 clock greater than the tWTR and tWR set-
ting normally required for the applicable speed bin to be JEDEC compliant; however,
Micron's DDR4 DRAMs do not require these additional tWTR and tWR clocks. The
CAS_n-to-CAS_n command delay to either a different bank group (tCCD_S) or the same
bank group (tCCD_L) have minimum timing requirements that must be satisfied be-
tween WRITE commands and are stated in the Timing Parameters by Speed Bin tables.
Figure 107: 1tCK vs. 2tCK WRITE Preamble Mode, tCCD = 4
1t CK Mode
CMD WRITE WRITE
CK_c
CK_t
tCCD =4 WL
DQS_t,
DQS_c Preamble
DQ D0 D1 D2 D3 D4 D5 D6 D7 D0 D1 D2 D3
2t CK Mode
CMD WRITE WRITE
CK_c
CK_t
tCCD =4 WL
DQS_t,
DQS_c Preamble
DQ D0 D1 D2 D3 D4 D5 D6 D7 D0 D1
Figure 108: 1tCK vs. 2tCK WRITE Preamble Mode, tCCD = 5
1t CK Mode
CMD WRITE WRITE
CK_c
CK_t
tCCD =5 WL
DQS_t,
DQS_c Preamble
Preamble
DQ D0 D1 D2 D3 D4 D5 D6 D7 D0 D1 D2 D3
2t CK Mode: t CCD = 5 is not allowed in 2t CK mode.
Note: 1. tCCD_S and tCCD_L = 5 tCKs is not allowed when in 2tCK WRITE preamble mode.
Figure 109: 1tCK vs. 2 tCK WRITE Preamble Mode, tCCD = 6
1t CK Mode
CMD WRITE WRITE
CK_c
CK_t
tCCD =6 WL
DQS_t,
DQS_c Preamble
Preamble
DQ D0 D1 D2 D3 D4 D5 D6 D7 D0 D1 D2 D3
2t CK Mode
CMD WRITE WRITE
CK_c
CK_t
tCCD =6 WL
DQS_t,
DQS_c Preamble
Preamble
DQ D0 D1 D2 D3 D4 D5 D6 D7 D0 D1 D2 D3
READ Preamble Mode

MR4[11] = 0 selects 1tCK READ preamble mode and MR4[11] = 1 selects 2tCK READ pre-
amble mode. Examples are shown in the following figure.
Figure 110: 1tCK vs. 2tCK READ Preamble Mode
1tCK Mode
RD
CL
CK_c
CK_t
Preamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
2tCK Mode
RD
CL
CK_c
CK_t
Preamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
READ Preamble Training

DDR4 supports READ preamble training via MPR reads; that is, READ preamble train-
ing is allowed only when the DRAM is in the MPR access mode. The READ preamble
training mode can be used by the DRAM controller to train or "read level" its DQS re-
ceivers. READ preamble training is entered via an MRS command (MR4[10] = 1 is ena-
bled and MR4[10] = 0 is disabled). After the MRS command is issued to enable READ
preamble training, the DRAM DQS signals are driven to a valid level by the time tSDO is
satisfied. During this time, the data bus DQ signals are held quiet, that is, driven HIGH.
The DQS_t signal remains driven LOW and the DQS_c signal remains driven HIGH until
an MPR Page0 READ command is issued (MPR0 through MPR3 determine which pat-
tern is used), and when CAS latency (CL) has expired, the DQS signals will toggle nor-
mally depending on the burst length setting. To exit READ preamble training mode, an
MRS command must be issued, MR4[10] = 0.
Figure 111: READ Preamble Training
CMD MRS MPR RD
tSDO CL
DQS_c
DQS_t,
DQs (Quiet/Driven HIGH) D0 D1 D2 D3 D4 D5 D6 D7
WRITE Postamble
Whether the 1tCK or 2tCK WRITE preamble mode is selected, the WRITE postamble re-
mains the same at ½tCK.
Figure 112: WRITE Postamble
1tCK Mode
WR
WL
CK_c
CK_t
Postamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
2tCK Mode
WR
WL
CK_c
CK_t
Postamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
READ Postamble
Whether the 1tCK or 2tCK READ preamble mode is selected, the READ postamble re-
mains the same at ½tCK.
Figure 113: READ Postamble
1tCK Mode
RD
CL
CK_c
CK_t
Postamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
2tCK Mode
RD
CL
CK_c
CK_t
Postamble
DQS_t,
DQS_c
DQ D0 D1 D2 D3 D4 D5 D6 D7
Bank Access Operation

DDR4 supports bank grouping: x4/x8 DRAMs have four bank groups (BG[1:0]), and
each bank group is comprised of four subbanks (BA[1:0]); x16 DRAMs have two bank
groups (BG[0]), and each bank group is comprised of four subbanks. Bank accesses to
different banks' groups require less time delay between accesses than bank accesses to
within the same bank's group. Bank accesses to different bank groups require tCCD_S
(or short) delay between commands while bank accesses within the same bank group
require tCCD_L (or long) delay between commands.
Figure 114: Bank Group x4/x8 Block Diagram

Bank 3 Bank 3 Bank 3 Bank 3
Memory Array Memory Array Memory Array Memory Array
Bank Group 0 Bank Group 1 Bank Group 2 Bank Group 3

CMD/ADDR CMD/ADDR
register
Sense amplifiers Sense amplifiers Sense amplifiers Sense amplifiers
Local I/O gating Local I/O gating Local I/O gating Local I/O gating
Global I/O gating
Data I/O
Notes: 1. Bank accesses to different bank groups require tCCD_S.

2. Bank accesses within the same bank group require tCCD_L.
Table 68: DDR4 Bank Group Timing Examples
Parameter DDR4-1600 DDR4-2133 DDR4-2400

tCCD_S 4nCK 4nCK 4nCK
tCCD_L 4nCK or 6.25ns 4nCK or 5.355ns 4nCK or 5ns
tRRD_S (½K) 4nCK or 5ns 4nCK or 3.7ns 4nCK or 3.3ns

tRRD_L (½K) 4nCK or 6ns 4nCK or 5.3ns 4nCK or 4.9ns
tRRD_S (1K) 4nCK or 5ns 4nCK or 3.7ns 4nCK or 3.3ns

tRRD_L (1K) 4nCK or 6ns 4nCK or 5.3ns 4nCK or 4.9ns
tRRD_S (2K) 4nCK or 6ns 4nCK or 5.3ns 4nCK or 5.3ns

tRRD_L (2K) 4nCK or 7.5ns 4nCK or 6.4ns 4nCK or 6.4ns
Table 68: DDR4 Bank Group Timing Examples (Continued)
Parameter DDR4-1600 DDR4-2133 DDR4-2400

tWTR_S 2nCK or 2.5ns 2nCK or 2.5ns 2nCK or 2.5ns
tWTR_L 4nCK or 7.5ns 4nCK or 7.5ns 4nCK or 7.5ns
Notes: 1. Refer to Timing Tables for actual specification values, these values are shown for refer-
ence only and are not verified for accuracy.
2. Timings with both nCK and ns require both to be satisfied; that is, the larger time of the
two cases must be satisfied.
Figure 115: READ Burst tCCD_S and tCCD_L Examples

T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
CK_c
CK_t
Command READ DES DES DES READ DES DES DES DES DES READ DES
tCCD_S tCCD_L
Bank Group
BG a BG b BG b
(BG)
Bank Bank c Bank c Bank c
Address Col n Col n Col n
Don’t Care
Notes: 1. tCCD_S; CAS_n-to-CAS_n delay (short). Applies to consecutive CAS_n to different bank
groups (T0 to T4).
2. tCCD_L; CAS_n-to-CAS_n delay (long). Applies to consecutive CAS_n to the same bank
group (T4 to T10).
Figure 116: Write Burst tCCD_S and tCCD_L Examples

T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
CK_c
CK_t
Command WRITE DES DES DES WRITE DES DES DES DES DES WRITE DES
tCCD_S tCCD_L
Bank Group
BG a BG b BG b
(BG)
Bank Bank c Bank c Bank c
Address Coln Coln Coln
Don’t Care
Notes: 1. tCCD_S; CAS_n-to-CAS_n delay (short). Applies to consecutive CAS_n to different bank
groups (T0 to T4).
2. tCCD_L; CAS_n-to-CAS_n delay (long). Applies to consecutive CAS_n to the same bank
group (T4 to T10).
Figure 117: tRRD Timing

T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
CK_c
CK_t
Command ACT DES DES DES ACT DES DES DES DES DES ACT DES
tRRD_S tRRD_L
Bank
Group BG a BG b BG b
(BG)
Bank Bank c Bank c Bank d
Address Row n Row n Row n
Don’t Care
Notes: 1. tRRD_S; ACTIVATE-to-ACTIVATE command period (short); applies to consecutive ACTI-

VATE commands to different bank groups (T0 and T4).
2. tRRD_L; ACTIVATE-to-ACTIVATE command period (long); applies to consecutive ACTI-
VATE commands to the different banks in the same bank group (T4 and T10).
Figure 118: tWTR_S Timing (WRITE-to-READ, Different Bank Group, CRC and DM Disabled)
T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Tb0 Tb1
CK_c
CK_t
Command WRITE Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid READ Valid
tWTR_S
Bank BGa BGb

Group
Bank Bank c Bank c
Address Col n Col n

tWPRE
tWPST
DQS, DQS_c
DQ DI DI DI DI DI DI DI DI
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
WL RL
Note: 1. tWTR_S: delay from start of internal write transaction to internal READ command to a
different bank group.
Figure 119: tWTR_L Timing (WRITE-to-READ, Same Bank Group, CRC and DM Disabled)
T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Tb0 Tb1
CK_c
CK_t
Command WRITE Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid READ Valid
tWTR_L
Bank BGa BGa

Group
Bank Bank c Bank c
Address Col n Col n

tWPRE
tWPST
DQS, DQS_c
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
WL RL
Note: 1. tWTR_L: delay from start of internal write transaction to internal READ command to the
same bank group.
READ Operation
READ Operation
Read Timing Definitions
The read timings shown below are applicable in normal operation mode, that is, when
the DLL is enabled and locked.
Note: tDQSQ = both rising/falling edges of DQS; no tAC defined.
Rising data strobe edge parameters:
• tDQSCK (MIN)/(MAX) describes the allowed range for a rising data strobe edge rela-
tive to CK.
• tDQSCK is the actual position of a rising strobe edge relative to CK.
• tQSH describes the DQS differential output HIGH time.
• tDQSQ describes the latest valid transition of the associated DQ pins.
• tQH describes the earliest invalid transition of the associated DQ pins.
Falling data strobe edge parameters:
• tQSL describes the DQS differential output LOW time.
READ Operation
Figure 120: Read Timing Definition

CK_c
CK_t
tDQSCK (MIN) tDQSCK (MAX) tDQSCK (MIN) tDQSCK (MAX)
tDQSCKi tDQSCKi
tDQSCK Rising strobe Rising strobe

MAX region region
window window
tDQSCKi tDQSCKi

center region region
window window
tDQSCKi tDQSCKi

MIN region region
window window
tDQSCK tDQSCK
tQSH/DQS_c tQSH/DQS_t
DQS_c
DQS_t
tQH tQH
tDQSQ tDQSQ
Associated
DQ Pins
Table 69: Read-to-Write and Write-to-Read Command Intervals
Access Type Bank Group Timing Parameters Note

Read-to-Write, mini- Same CL - CWL + RBL/2 + 1tCK + tWPRE 1, 2
mum Different CL - CWL + RBL/2 + 1tCK + tWPRE 1, 2
Same CWL + WBL/2 + tWTR_L 1, 3
Write-to-Read, mini-
mum Different CWL + WBL/2 + tWTR_S 1, 3
Notes: 1. These timings require extended calibrations times tZQinit and tZQCS.
2. RBL: READ burst length associated with READ command, RBL = 8 for fixed 8 and on-the-
fly mode 8 and RBL = 4 for fixed BC4 and on-the-fly mode BC4.
3. WBL: WRITE burst length associated with WRITE command, WBL = 8 for fixed 8 and on-
the-fly mode 8 or BC4 and WBL = 4 for fixed BC4 only.
Read Timing – Clock-to-Data Strobe Relationship

The clock-to-data strobe relationship shown below is applicable in normal operation
mode, that is, when the DLL is enabled and locked.
READ Operation

• tDQSCK (MIN)/(MAX) describes the allowed range for a rising data strobe edge rela-
tive to CK.
t
• DQSCK is the actual position of a rising strobe edge relative to CK.
• tQSH describes the data strobe high pulse width.
• tHZ(DQS) DQS strobe going to high, nondrive level (shown in the postamble section
of the figure below).
• tQSL describes the data strobe low pulse width.
• tLZ(DQS) DQS strobe going to low, initial drive level (shown in the preamble section
of the figure below).
Figure 121: Clock-to-Data Strobe Relationship

RL measured
to this point
CK_t
CK_c
tDQSCK (MIN) tDQSCK (MIN) tDQSCK (MIN) tDQSCK (MIN)
tHZ(DQS) MIN
tLZ(DQS) MIN tQSH tQSL tQSH tQSL tQSH tQSL
DQS_t, DQS_c
Early Strobe Bit 0 Bit 1 Bit 2 Bit 3 Bit 4 Bit 5 Bit 6 Bit 7
tRPRE tRPST
tHZ(DQS) MAX
tDQSCK (MAX) tDQSCK (MAX) tDQSCK (MAX) tDQSCK (MAX)

tRPST
tLZ(DQS) MAX
DQS_t, DQS_c
Late Strobe Bit 0 Bit 1 Bit 2 Bit 3 Bit 4 Bit 5 Bit 6 Bit 7
tRPRE tQSH tQSL tQSH tQSL
Notes: 1. Within a burst, the rising strobe edge will vary within tDQSCKi while at the same volt-
age and temperature. However, when the device, voltage, and temperature variations
are incorporated, the rising strobe edge variance window can shift between tDQSCK
(MIN) and tDQSCK (MAX).
A timing of this window's right edge (latest) from rising CK_t, CK_c is limited by a devi-
ce's actual tDQSCK (MAX). A timing of this window's left inside edge (earliest) from ris-
ing CK_t, CK_c is limited by tDQSCK (MIN).
2. Notwithstanding Note 1, a rising strobe edge with tDQSCK (MAX) at T(n) can not be im-
mediately followed by a rising strobe edge with tDQSCK (MIN) at T(n + 1) because other
timing relationships (tQSH, tQSL) exist: if tDQSCK(n + 1) < 0: tDQSCK(n) < 1.0 tCK - (tQSH
(MIN) + tQSL (MIN)) - | tDQSCK(n + 1) |.
3. The DQS_t, DQS_c differential output HIGH time is defined by tQSH, and the DQS_t,
DQS_c differential output LOW time is defined by tQSL.
4. tLZ(DQS) MIN and tHZ(DQS) MIN are not tied to tDQSCK (MIN) (early strobe case), and
tLZ(DQS) MAX and tHZ(DQS) MAX are not tied to tDQSCK (MAX) (late strobe case).
5. The minimum pulse width of READ preamble is defined by tRPRE (MIN).

6. The maximum READ postamble is bound by tDQSCK (MIN) plus tQSH (MIN) on the left
side and tHZDSQ (MAX) on the right side.
7. The minimum pulse width of READ postamble is defined by tRPST (MIN).
READ Operation
8. The maximum READ preamble is bound by tLZDQS (MIN) on the left side and tDQSCK
(MAX) on the right side.
Read Timing – Data Strobe-to-Data Relationship

The data strobe-to-data relationship is shown below and is applied when the DLL is en-
abled and locked.
Note: tDQSQ: both rising/falling edges of DQS; no tAC defined.
Data valid window parameters:
• tDVWd is the Data Valid Window per device per UI and is derived from [ tQH - tDQSQ]
of each UI on a given DRAM
• tDVWp is the Data Valid Window per pin per UI and is derived [ tQH - tDQSQ] of each
UI on a pin of a given DRAM
Figure 122: Data Strobe-to-Data Relationship

T0 T1 T2 T9 T10 T11 T12 T13 T14 T15 T16
CK_c
CK_t
Command3 READ DES DES DES DES DES DES DES DES DES DES
RL = AL + CL
Bank,
Address4 Col n
tDQSQ (MAX) tDQSQ (MAX)
tRPST
tRPRE (1nCK)
DQS_t, DQS_c
tQH tQH
DQ2 DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT

(Last data ) n n+1 n+2 n+3 n+4 n+5 n+6 n+7
tDVWp
DQ2 DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT

n n+1 n+2 n+3 n+4 n+5 n+6 n+7
(First data no longer)
tDVWp
DOUT DOUT DOUT DOUT DOUT DOUT DOUT DOUT
All DQ collectively n n+1 n+2 n+3 n+4 n+5 n+6 n+7
tDVWd tDVWd
Don’t Care
Notes: 1. BL = 8, RL = 11 (AL = 0, CL = 1) , Premable = 1tCK.

2. DOUTn = data-out from column n.
3. DES commands are shown for ease of illustration; other commands may be valid at
these times.
4. BL8 setting activated by either MR0[A1:0 = 00] or MR0[A1:0 = 01] and A12 = 1 during
READ commands at T0.
5. Output timings are referenced to VDDQ, and DLL on for locking.
6. tDQSQ defines the skew between DQS to data and does not define DQS to clock.
READ Operation
7. Early data transitions may not always happen at the same DQ. Data transitions of a DQ
can vary (either early or late) within a burst.
tLZ(DQS), tLZ(DQ), tHZ(DQS), and tHZ(DQ) Calculations

tHZ and tLZ transitions occur in the same time window as valid data transitions. These
parameters are referenced to a specific voltage level that specifies when the device out-
put is no longer driving tHZ(DQS) and tHZ(DQ), or begins driving tLZ(DQS) and
tLZ(DQ). The figure below shows a method to calculate the point when the device is no
longer driving tHZ(DQS) and tHZ(DQ), or begins driving tLZ(DQS) and tLZ(DQ), by
measuring the signal at two different voltages. The actual voltage measurement points
are not critical as long as the calculation is consistent. tLZ(DQS), tLZ(DQ), tHZ(DQS),
and tHZ(DQ) are defined as singled-ended parameters.
Figure 123: tLZ and tHZ Method for Calculating Transitions and Endpoints
tLZ(DQ): CK_t, CK_c rising crossing at RL tHZ(DQ) with BL8: CK_t, CK_c rising crossing at RL + 4CK
tHZ(DQ) with BC4: CK_t, CK_c rising crossing at RL + 2CK
CK_t
CK_c
tLZ tHZ
Begin point:
Extrapolated point at VDDQ
VDDQ
DQ VDDQ DQ
VSW2 VSW2
0.7 × VDDQ 0.7 × VDDQ
VSW1 VSW1
0.4 × VDDQ 0.4 × VDDQ

Begin point: Extrapolated point (low level)
Notes: 1. Vsw1 = (0.70 - 0.04) × VDDQ for both tLZ and tHZ.
2. Vsw2 = (0.70 + 0.04) × VDDQ for both tLZ and tHZ.
3. Extrapolated point (low level) = VDDQ/(50 + 34) × 34 = 0.4 × VDDQ
Driver impedance = RZQ/7 = 34˖
VTT test load = 50˖ to VDDQ.
READ Operation
tRPRE Calculation
Figure 124: tRPRE Method for Calculating Transitions and Endpoints
CK_t
VDD /2
CK_c
Single-ended signal provided as background information

DQS_t VDDQ
0.7 × VDDQ
0.4 × VDDQ
DQS_c
VDDQ
0.7 × VDDQ
0.4 × VDDQ
DQS_t DQS_t VDDQ

DQS_c
0.7 × VDDQ
DQS_c 0.4 × VDDQ
Resulting differential signal relevant for tRPRE specification

0.6 × VDDQ
VSW2
0.3 × VDDQ
VSW1
DQS_t, DQS_c 0V
t
RPRE begins (t1) t
RPRE ends (t2)
Notes: 1. Vsw1 = (0.3 - 0.04) × VDDQ.

2. Vsw2 = (0.30 + 0.04) × VDDQ.
3. DQS_t and DQS_c low level = VDDQ/(50 + 34) × 34 = 0.4 × VDDQ
READ Operation
tRPST Calculation
Figure 125: tRPST Method for Calculating Transitions and Endpoints
CK_t
VDD /2
CK_c
Single-ended signal provided as background information

VDDQ
0.7 × VDDQ
DQS_t 0.4 × VDDQ

DQS_c VDDQ
0.7 × VDDQ
0.4 × VDDQ
DQS_c VDDQ
0.7 × VDDQ
DQS_t
Resulting differential signal relevant fortRPST specification
tRPST beginst(1)
0V
VSW2
–0.3 × VDDQ
VSW1
DQS_t, DQS_c –0.6 × VDDQ

tRPST ends t(2)
Notes: 1. Vsw1 = (–0.3 - 0.04) × VDDQ.

2. Vsw2 = (–0.30 + 0.04) × VDDQ.
3. DQS_t and DQS_c low level = VDDQ/(50 + 34) × 34 = 0.4 × VDDQ
READ Operation
READ Burst Operation

DDR4 READ commands support bursts of BL8 (fixed), BC4 (fixed), and BL8/BC4 on-
the-fly (OTF); OTF uses address A12 to control OTF when OTF is enabled:
• A12 = 0, BC4 (BC4 = burst chop)
• A12 = 1, BL8
READ commands can issue precharge automatically with a READ with auto precharge
command (RDA), and is enabled by A10 HIGH:
• READ command with A10 = 0 (RD) performs standard read, bank remains active after
READ burst.
• READ command with A10 = 1 (RDA) performs read with auto precharge, bank goes in
to precharge after READ burst.
Figure 126: READ Burst Operation, RL = 11 (AL = 0, CL = 11, BL8)

T0 T1 T2 Ta0 Ta1 Ta2 Ta3 Ta4 Ta5 Ta6 Ta7 Ta8 Ta9
CK_c
CK_t
Command READ DES DES DES DES DES DES DES DES DES DES DES DES
Bank Group
BGa
Address
Address Bank
col n
tRPRE tRPST
DQS_t
DQS_c
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
CL = 11
RL = AL + CL
Notes: 1. BL8, RL = 0, AL = 0, CL = 11, Preamble = 1tCK.

2. DO n = data-out from column n.
these times.
4. BL8 setting activated by either MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during READ
command at T0.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable.
READ Operation
Figure 127: READ Burst Operation, RL = 21 (AL = 10, CL = 11, BL8)

T0 T1 Ta0 Ta1 Ta2 Ta3 Tb0 Tb1 Tb2 Tb3 Tb4 Tb5 Tb6
CK_c
CK_t
Command READ DES DES DES DES DES DES DES DES DES DES DES DES
Bank Group
BGa
Address
Address Bank
col n
tRPRE tRPST
DQS_t
DQS_c
n n+ 1 n+ 2 n+ 3 n+ 4 n+ 5 n+ 6 n+ 7
AL = 10 CL = 11
RL = AL + CL
Notes: 1. BL8, RL = 21, AL = (CL - 1), CL = 11, Preamble = 1tCK.

these times.
command at T0.
READ Operation
READ Operation Followed by Another READ Operation
Figure 128: Consecutive READ (BL8) with 1tCK Preamble in Different Bank Group
CK_c
CK_t
Command READ DES DES DES READ DES DES DES DES DES DES DES DES DES DES DES DES DES
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DO DO DO DO DO DO DO DO DO DO DO DO
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
RL = 11
Notes: 1. BL8, AL = 0, CL = 11, Preamble = 1tCK.

2. DO n (or b) = data-out from column n (or column b).
these times.
commands at T0 and T4.
Figure 129: Consecutive READ (BL8) with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
RL = 11

these times.
READ Operation
Figure 130: Nonconsecutive READ (BL8) with 1tCK Preamble in Same or Different Bank Group
CK_c
CK_t
Command READ DES DES DES DES READ DES DES DES DES DES DES DES DES DES DES DES DES
tCCD_S/L =5
Bank Group
BGa BGb
Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
RL = 11
Notes: 1. BL8, AL = 0, CL = 11, Preamble = 1tCK, tCCD_S/L = 5.

these times.
Figure 131: Nonconsecutive READ (BL8) with 2tCK Preamble in Same or Different Bank Group
CK_c
CK_t
tCCD_S/L =6
Bank Group BGa BGa or

Address BGb
Address Bank Bank

Col n Col b
tRPRE tRPRE tRPST
DQS_t,
DQS_c
RL = 11
RL = 11
Notes: 1. BL8, AL = 0, CL = 11, Preamble = 2tCK, tCCD_S/L = 6.

these times.
4. BL8 setting activated by either MR0[A1:0 = 00] or MR0[A1:0 = 01] and A12 = 1 during
READ commands at T0 and T6.
6. 6 tCCD_S/L = 5 isn’t allowed in 2tCK preamble mode.
READ Operation
Figure 132: READ (BC4) to READ (BC4) with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
RL = 11

these times.
4. BC4 setting activated by either MR0[1:0] = 10 or MR0[1:0] = 01 and A12 = 0 during READ
Figure 133: READ (BC4) to READ (BC4) with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
RL = 11

these times.
4. BC4 setting activated by either MR0[1:0] = 10 or MR0[1:0] = 01 and A12 = 0 during READ
READ Operation
Figure 134: READ (BL8) to READ (BC4) OTF with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
t RPRE tRPST
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DO DO DO DO DO DO DO DO
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
RL = 11
Notes: 1. BL = 8, AL = 0, CL = 11, Preamble = 1tCK.

these times.
4. BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T0. BC4
setting activated by MR0[1:0] = 01 and A12 = 0 during READ commands at T4.
Figure 135: READ (BL8) to READ (BC4) OTF with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
RL = 11
Notes: 1. BL = 8, AL =0, CL = 11, Preamble = 2tCK.

these times.
4. BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T0. BC4
READ Operation
Figure 136: READ (BC4) to READ (BL8) OTF with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
RL = 11
Notes: 1. BL = 8, AL =0, CL = 11, Preamble = 1tCK.

these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during READ commands at T0. BL8
Figure 137: READ (BC4) to READ (BL8) OTF with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
RL = 11
Notes: 1. BL = 8, AL = 0, CL = 11, Preamble = 2tCK.

these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during READ commands at T0. BL8
READ Operation
READ Operation Followed by WRITE Operation
Figure 138: READ (BL8) to WRITE (BL8) with 1tCK Preamble in Same or Different Bank Group
CK_c
CK_t
Command READ DES DES WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES DES
tWR
READ to WRITE command delay tWTR

= RL +BL/2 - WL + 2 tCK 4 Clocks
Bank Group BGa or

BGa BGb
Address
Address Bank Bank

Col n Col b
tRPRE tRPST tWPRE tWPST
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DO DO DO DO DI DI DI DI DI DI DI DI
WL = 9
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK, WL = 9 (CWL = 9, AL = 0), WRITE
preamble = 1tCK.
2. DO n = data-out from column n; DI b = data-in from column b.
these times.
commands at T0 and WRITE commands at T8.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, Write DBI = Disable,
Write CRC = Disable.
Figure 139: READ (BL8) to WRITE (BL8) with 2tCK Preamble in Same or Different Bank Group
CK_c
CK_t
t
WR
READ to WRITE command delay
= RL +BL/2 - WL + 3 tCK 4 Clocks t
WTR
Bank Group BGa or

BGa BGb
Address
Address Bank Bank

Col n Col b
t
t
RPRE t
RPST t
WPRE WPST
DQS_t,
DQS_c
RL = 11
WL = 10
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 2tCK, WL = 10 (CWL = 9+1 [see Note
5], AL = 0), WRITE preamble = 2tCK.
these times.
READ Operation
5. When operating in 2tCK WRITE preamble mode, CWL may need to be programmed to a
value at least 1 clock greater than the lowest CWL setting.
Figure 140: READ (BC4) OTF to WRITE (BC4) OTF with 1tCK Preamble in Same or Different Bank
Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DI DI DI DI
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = 9
Notes: 1. BC = 4, RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK, WL = 9 (CWL = 9, AL = 0),

WRITE preamble = 1tCK.
these times.
4. BC4 (OTF) setting activated by MR0[1:0] = 01 and A12 = 0 during READ commands at T0
and WRITE commands at T6.
READ Operation
Figure 141: READ (BC4) OTF to WRITE (BC4) OTF with 2tCK Preamble in Same or Different Bank
Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = 10
Notes: 1. BC = 4, RL = 11 (CL = 11, AL = 0), READ preamble = 2tCK, WL = 10 (CWL = 9 + 1 [see Note
these times.
4. BC4 (OTF) setting activated by MR0[1:0] = 01 and A12 = 0 during READ commands at T0
and WRITE commands at T6.
Figure 142: READ (BC4) Fixed to WRITE (BC4) Fixed with 1tCK Preamble in Same or Different Bank
Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = 9
Notes: 1. BC = 4, RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK, WL = 9 (CWL = 9, AL = 0),

WRITE preamble = 1tCK.
READ Operation
these times.
4. BC4 (fixed) setting activated by MR0[1:0] = 01.
Figure 143: READ (BC4) Fixed to WRITE (BC4) Fixed with 2tCK Preamble in Same or Different Bank
Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = 10
Notes: 1. BC = 4, RL = 11 (CL = 11, AL = 0), READ preamble = 2tCK, WL = 9 (CWL = 9 + 1 [see Note
these times.
READ Operation
Figure 144: READ (BC4) to WRITE (BL8) OTF with 1tCK Preamble in Same or Different Bank Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
tRPST tWPST
tRPRE tWPRE
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DI DI DI DI DI DI DI DI
n n+1 n+2 n+3 b b+1 b+2 b+3 n+4 n+5 n+6 n+7
WL = 9
preamble = 1tCK.
these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE commands at T0.
BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T6.
Figure 145: READ (BC4) to WRITE (BL8) OTF with 2tCK Preamble in Same or Different Bank Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
tRPST tWPST
tRPRE tWPRE
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DI DI DI DI DI DI DI DI
n n+1 n+2 n+3 b b+1 b+2 b+3 n+4 n+5 n+6 n+7
WL = 10
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 2tCK, WL = 10 (CWL = 9 + 1 [see Note
these times.
BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T6.
READ Operation
Figure 146: READ (BL8) to WRITE (BC4) OTF with 1tCK Preamble in Same or Different Bank Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DO DO DO DO DI DI DI DI
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
WL = 9
preamble = 1tCK.
these times.
4. BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T0.
BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE commands at T8.
Figure 147: READ (BL8) to WRITE (BC4) OTF with 2tCK Preamble in Same or Different Bank Group
CK_c
CK_t
tWR


Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
DQ DO DO DO DO DO DO DO DO DI DI DI DI
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
WL = 10
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 2tCK, WL = 10 (CWL = 9 + 1 [see Note
these times.
READ Operation
4. BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during READ commands at T0.
BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE commands at T8.
READ Operation Followed by PRECHARGE Operation

The minimum external READ command to PRECHARGE command spacing to the same
bank is equal to AL + tRTP with tRTP being the internal READ command to PRECHARGE
command delay. Note that the minimum ACT to PRE timing, tRAS, must be satisfied as
well. The minimum value for the internal READ command to PRECHARGE command
delay is given by tRTP (MIN) = MAX (4 × nCK, 7.5ns). A new bank ACTIVATE command
may be issued to the same bank if the following two conditions are satisfied simultane-
ously:
• The minimum RAS precharge time (tRP [MIN]) has been satisfied from the clock at
which the precharge begins.
• The minimum RAS cycle time (tRC [MIN]) from the previous bank activation has been
satisfied.
Figure 148: READ to PRECHARGE with 1tCK Preamble

CK_c
CK_t
Command DES READ DES DES DES PRE DES DES DES DES DES DES DES DES ACT DES DES DES
Bank Group BGa

BGa or BGb BGa
Address
Address Bank a Bank a Bank a

Col n (or all) Row b
tRTP tRP
RL = AL + CL
BC4 Opertaion
DQS_t,
DQS_c
DQ DO DO DO DO
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. RL = 11 (CL = 11, AL = 0 ), Preamble = 1tCK, tRTP = 6, tRP = 11.

these times.
4. The example assumes that tRAS (MIN) is satisfied at the PRECHARGE command time (T7)
and that tRC (MIN) is satisfied at the next ACTIVATE command time (T18).
READ Operation
Figure 149: READ to PRECHARGE with 2tCK Preamble

CK_c
CK_t
Command DES READ DES DES DES PRE DES DES DES DES DES DES DES DES ACT DES DES DES
Bank Group BGa or

BGa BGb BGa
Address

tRTP tRP
RL = AL + CL
BC4 Opertaion
DQS_t,
DQS_c
DQ DO DO DO DO
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7

these times.
4. The example assumes that tRAS (MIN) is satisfied at the PRECHARGE command time (T7)
and that tRC (MIN) is satisfied at the next ACTIVATE command time (T18).
Figure 150: READ to PRECHARGE with Additive Latency and 1tCK Preamble
CK_c
CK_t
Command DES READ DES DES DES DES DES DES PRE DES DES DES DES DES DES DES DES ACT
Bank Group BGa or

BGa BGb BGa
Address

AL = CL - 2 = 9 tRTP tRP
CL = 11
BC4 Opertaion
DQS_t,
DQS_c
DQ DO DO DO DO
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. RL =20 (CL = 11, AL = CL - 2), Preamble = 1tCK, tRTP = 6, tRP = 11.
READ Operation
these times.
4. The example assumes that tRAS (MIN) is satisfied at the PRECHARGE command time
(T16) and that tRC (MIN) is satisfied at the next ACTIVATE command time (T27).
Figure 151: READ with Auto Precharge and 1tCK Preamble

CK_c
CK_t
Command DES RDA DES DES DES PRE DES DES DES DES DES DES DES DES ACT DES DES DES
Bank Group BGa or

BGa BGb BGa
Address

Col n Col n Row b
tRTP tRP
RL = AL + CL
BC4 Opertaion
DQS_t,
DQS_c
DQ DO DO DO DO
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7

these times.
4. tRTP = 6 setting activated by MR0[A11:9 = 001].
5. The example assumes that tRC (MIN) is satisfied at the next ACTIVATE command time
(T18).
READ Operation
Figure 152: READ with Auto Precharge, Additive Latency, and 1tCK Preamble
CK_c
CK_t
Command DES RDA DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES ACT
Bank Group
BGa BGa
Address
Address Bank a Bank a

Col n Row b
AL = CL - 2 = 9 tRTP tRP
CL = 11
BC4 Opertaion
DQS_t,
DQS_c
DQ DO DO DO DO
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. RL = 20 (CL = 11, AL = CL - 2), Preamble = 1tCK, tRTP = 6, tRP = 11.

these times.
4. tRTP = 6 setting activated by MR0[11:9] = 001.
5. The example assumes that tRC (MIN) is satisfied at the next ACTIVATE command time
(T27).
READ Operation with Read Data Bus Inversion (DBI)
Figure 153: Consecutive READ (BL8) with 1tCK Preamble and DBI in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11 + 2 (Read DBI adder)
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b + 3 b +4 _ b + 5 b+6 b+7
RL = 11 + 2 (Read DBI adder)
DBI_n DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI DBI
Notes: 1. BL = 8, AL = 0, CL = 11, Preamble = 1tCK, RL = 11 + 2 (Read DBI adder).
READ Operation
2. DO n (or b) = data-out from column n (or b); DBI n (or b) = data bus inversion from col-
umn n (or b).
these times.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Enable.
READ Operation with Command/Address Parity (CA Parity)
Figure 154: Consecutive READ (BL8) with 1tCK Preamble and CA Parity in Different Bank Group
CK_c
CK_t
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
Parity
tRPRE tRPST
DQS_t,
DQS_c
RL = 15
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b + 3 b +4 _ b + 5 b+6 b+7
RL = 15
Notes: 1. BL = 8, AL = 0, CL = 11, PL = 4, (RL = CL + AL + PL = 15), Preamble = 1tCK.

2. DO n (or b) = data-out from column n (or b).
these times.
4. BL8 setting activated by either MR0[A1:A0 = 00] or MR0[A1:A0 = 01] and A12 = 1 during
READ commands at T0 and T4.
5. CA parity = Enable, CS to CA latency = Disable, Read DBI = Disable.
READ Operation
Figure 155: READ (BL8) to WRITE (BL8) with 1tCK Preamble and CA Parity in Same or Different Bank
Group
CK_c
CK_t
tWR
= RL +BL/2 - WL + 2 tCK 4 Clocks tWTR

Address BGb
Address Bank Bank

Col n Col b
Parity
DQS_t,
DQS_c
RL = 15
WL = 13
Notes: 1. BL = 8, AL = 0, CL = 11, PL = 4, (RL = CL + AL + PL = 15), READ preamble = 1tCK, CWL = 9,

AL = 0, PL = 4, (WL = CL + AL + PL = 13), WRITE preamble = 1tCK.
2. DO n = data-out from column n, DI b = data-in from column b.
these times.
commands at T0 and WRITE command at T8.
5. CA parity = Enable, CS to CA latency = Disable, Read DBI = Disable, Write DBI = Disable,
READ Operation
READ Followed by WRITE with CRC Enabled
Figure 156: READ (BL8) to WRITE (BL8 or BC4: OTF) with 1tCK Preamble and Write CRC in Same or
Different Bank Group
CK_c
CK_t
tWR
= RL +BL/2 - WL + 2 tCK 4 Clocks tWTR

Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
DQ x4, DO DO DO DO DO DO DO DO DI DI DI DI DI DI DI DI CRC CRC

BL = 8
WL = 9
DQ x8/X16, DO DO DO DO DO DO DO DO DI DI DI DI DI DI DI DI CRC
BL = 8
DQ x4, DO DO DO DO DO DO DO DO DI DI DI DI
READ: BL = 8, n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3 CRC CRC
WRITE: BC = 4 (OTF)
DQ x8/X16,
DO DO DO DO DO DO DO DO DI DI DI DI CRC
READ: BL = 8, n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
WRITE: BC = 4 (OTF)
Notes: 1. BL = 8 (or BC = 4: OTF for Write), RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK, WL =
9 (CWL = 9, AL = 0), WRITE preamble = 1tCK.
these times.
Write CRC = Enable.
READ Operation
Figure 157: READ (BC4: Fixed) to WRITE (BC4: Fixed) with 1tCK Preamble and Write CRC in Same or
Different Bank Group
CK_c
CK_t
tWR
2 Clocks tWTR
= RL +BL/2 - WL + 2 tCK

Address BGb
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
RL = 11
DQ x4, DO DO DO DO DI DI DI DI CRC CRC

n n+1 n+2 n+3 b b+1 b+2 b+3
BC = 4 (Fixed)
WL = 9
DQ x8/X16, DO DO DO DO DI DI DI DI CRC
n n+1 n+2 n+3 b b+1 b+2 b+3
BC = 4 (Fixed)
Notes: 1. BC = 4 (Fixed), RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK, WL = 9 (CWL = 9, AL =

0), WRITE preamble = 1tCK.
these times.
4. BC4 setting activated by MR0[1:0] = 10.
Write CRC = Enable.
READ Operation with Command/Address Latency (CAL) Enabled
Figure 158: Consecutive READ (BL8) with CAL (3tCK) and 1tCK Preamble in Different Bank Group
CK_c
CK_t
tCAL =3 tCAL =3
Command DES DES READ DES DES DES READ DES DES DES DES DES DES DES DES DES DES
w/o CS_n
CS_n
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
DQ DI DI DI DI DI DI DI DI DI DI DI DI
n n+1 n+2 n+5 n+6 n+7 b b+1 b+2 b+5 b+6 b+7
RL = 11
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK.
READ Operation
2. DI n (or b) = data-in from column n (or b).

these times.
5. CA parity = Disable, CS to CA latency = Enable, Read DBI = Disable, Write DBI = Disable,
6. Enabling CAL mode does not impact ODT control timings. The same timing relationship
relative to the command/address bus as when CAL is disabled should be maintained.
Figure 159: Consecutive READ (BL8) with CAL (4tCK) and 1tCK Preamble in Different Bank Group
CK_c
CK_t
tCAL =4 tCAL =4
Command DES DES READ DES DES DES READ DES DES DES DES DES DES DES DES DES DES
w/o CS_n
CS_n
tCCD_S =4
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tRPRE tRPST
DQS_t,
DQS_c
RL = 11
n n+1 n+2 n+5 n+6 n+7 b b+1 b+2 b+5 b+6 b+7
RL = 11
Notes: 1. BL = 8, RL = 11 (CL = 11, AL = 0), READ preamble = 1tCK.

2. DI n (or b) = data-in from column n (or b).
these times.
5. CA parity = Disable, CS to CA latency = Enable, Read DBI = Disable, Write DBI = Disable,
6. Enabling CAL mode does not impact ODT control timings. The same timing relationship
relative to the command/address bus as when CAL is disabled should be maintained.
WRITE Operation
WRITE Operation
Write Timing Definitions
The write timings shown in the following figures are applicable in normal operation
mode, that is, when the DLL is enabled and locked.
Write Timing – Clock-to-Data Strobe Relationship

The clock-to-data strobe relationship is shown below and is applicable in normal oper-
ation mode, that is, when the DLL is enabled and locked.
• tDQSS (MIN) to tDQSS (MAX) describes the allowed range for a rising data strobe edge
relative to CK.
• tDQSS is the actual position of a rising strobe edge relative to CK.
• tDQSH describes the data strobe high pulse width.
• tWPST strobe going to HIGH, nondrive level (shown in the postamble section of the
graphic below).
• tDQSL describes the data strobe low pulse width.
• tWPRE strobe going to LOW, initial drive level (shown in the preamble section of the
graphic below).
WRITE Operation
Figure 160: Write Timing Definition

T0 T1 T2 T7 T8 T9 T10 T11 T12 T13 T14
CK_c
CK_t
Command3 WRITE DES DES DES DES DES DES DES DES DES DES
WL = AL + CWL
Address4 Bank,
Col n
tDQSS tDSH tDSH tDSH tDSH
tDQSS tWPRE(1nCK) tWPSTaa

(MIN)
DQS_t, DQS_c
tDQSL tDQSH tDQSL tDQSH tDQSL tDQSH tDQSL tDQSH
tDQSH (MIN) tDQSL (MIN)
tDSS tDSS tDSS tDSS tDSS
DQ2 DIN
n
DIN
n+ 2
DIN
n+ 3
DIN
n+ 4
DIN
n+ 6
DIN
n+ 7
tDSH tDSH tDSH tDSH

tDQSS tWPRE(1nCK) tWPST (MIN)
(nominal)
DQS_t, DQS_c
tDQSH tDQSL (MIN)
(MIN)
DQ2 DIN
n
DIN
n+ 2
DIN
n+ 3
DIN
n+ 4
DIN
n+ 6
DIN
n+ 7
tDQSS
tDSH tDSH tDSH tDSH

tWPRE(1nCK) tWPST (MIN)
tDQSS (MAX)
DQS_t, DQS_c
tDQSH tDQSL (MIN)

(MIN)
DIN DIN DIN DIN DIN DIN

DQ2 n n+ 2 n+ 3 n+ 4 n+ 6 n+ 7
DM_n
Notes: 1. BL8, WL = 9 (AL = 0, CWL = 9).

2. DINn = data-in from column n.
these times.
4. BL8 setting activated by either MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during
WRITE command at T0.
5. tDQSS must be met at each rising clock edge.
WRITE Operation
tWPRE Calculation
Figure 161: tWPRE Method for Calculating Transitions and Endpoints
&.BW
9''
&.BF
6LQJOHHQGHGVLJQDOSURYLGHGDVEDFNJURXQGLQIRUPDWLRQ
'46BW
95()'4

'46BF
95()'4

'46BW '46BW
'46BF

95()'4
'46BF
5HVXOWLQJGLIIHUHQWLDOVLJQDOUHOHYDQWIRUW :35(VSHFLILFDWLRQ
9,+',))3HDN
9,+',))'46

96:
96:
'46BW'46BF 9
W :35(EHJLQV W W :35(HQGV W
Notes: 1. Vsw1 = (0.1) × VIH,diff,DQS.

2. Vsw2 = (0.9) × VIH,diff,DQS.
WRITE Operation
tWPST Calculation
Figure 162: tWPST Method for Calculating Transitions and Endpoints
&.BW
9''
&.BF
6LQJOHHQGHGVLJQDOSURYLGHGDVEDFNJURXQGLQIRUPDWLRQ

95()'4
'46BW
'46BF

95()'4
'46BF
95()'4

'46BW
5HVXOWLQJGLIIHUHQWLDOVLJQDOUHOHYDQWIRU W :367VSHFLILFDWLRQ
W :367EHJLQV W
9
96:
96:

9,/',))'46
'46BW'46BF 9,/',))3HDN
W :367HQGV W
Notes: 1. Vsw1 =(0.9) × VIL,diff,DQS.

2. Vsw2 = (0.1) × VIL,diff,DQS.
Write Timing – Data Strobe-to-Data Relationship

The DQ input receiver uses a compliance mask (Rx) for voltage and timing as shown in
the figure below. The receiver mask (Rx mask) defines the area where the input signal
must not encroach in order for the DRAM input receiver to be able to successfully cap-
ture a valid input signal. The Rx mask is not the valid data-eye. TdiVW and V diVW define
the absolute maximum Rx mask.
WRITE Operation
Figure 163: Rx Compliance Mask
Rx Mask
VDIVW VCENTDQ,midpoint
TdiVW
VCENTDQ,midpoint is defined as the midpoint between the largest V REFDQ voltage level and
the smallest V REFDQ voltage level across all DQ pins for a given DRAM. Each DQ pin's
VREFDQ is defined by the center (widest opening) of the cumulative data input eye as de-
picted in the following figure. This means a DRAM's level variation is accounted for
within the DRAM Rx mask. The DRAM V REFDQ level will be set by the system to account
for RON and ODT settings.
Figure 164: VCENT_DQ VREFDQ Voltage Variation
DQx DQy DQz

(smallest VREFDQ Level) (largest VREFDQ Level)
VCENTDQx VCENTDQz
VCENTDQ,midpoint
VCENTDQy
VREF variation
(component)
The following figure shows the Rx mask requirements both from a midpoint-to-mid-
point reference (left side) and from an edge-to-edge reference. The intent is not to add
any new requirement or specification between the two but rather how to convert the
relationship between the two methodologies. The minimum data-eye shown in the
composite view is not actually obtainable due to the minimum pulse width require-
ment.
WRITE Operation
Figure 165: Rx Mask DQ-to-DQS Timings
DQS, DQs Data-In at DRAM Ball DQS, DQs Data-In at DRAM Ball
Rx Mask Rx Mask – Alternative View
DQS_c DQS_c
DQS_t DQS_t
0.5 × TdiVW 0.5 × TdiVW 0.5 × TdiVW 0.5 × TdiVW
DRAMa DRAMa
VdiVW
VdiVW
DQx–z Rx Mask DQx–z Rx Mask
TdiVW TdiVW
tDQS2DQ +0.5 × TdiVW

tDQS2DQ
DRAMb DRAMb Rx Mask
VdiVW
VdiVW
DQy
Rx Mask DQy TdiVW
tDQ2DQ tDQ2DQ
DRAMb DRAMb Rx Mask
VdiVW
VdiVW
DQz
Rx Mask DQz TdiVW
tDQ2DQ
tDQS2DQ +0.5 × TdiVW

tDQS2DQ
DRAMc DRAMc Rx Mask

VdiVW
VdiVW
DQz
Rx Mask DQz TdiVW
tDQ2DQ tDQ2DQ
DRAMc DRAMc Rx Mask

VdiVW
VdiVW
DQy
Rx Mask DQy TdiVW
tDQ2DQ
Notes: 1. DQx represents an optimally centered mask.

DQy represents earliest valid mask.
DQz represents latest valid mask.
2. DRAMa represents a DRAM without any DQS/DQ skews.
DRAMb represents a DRAM with early skews (negative tDQS2DQ).
DRAMc represents a DRAM with delayed skews (positive tDQS2DQ).
3. This figure shows the skew allowed between DRAM-to-DRAM and between DQ-to-DQ
for a DRAM. Signals assume data is center-aligned at DRAM latch.
TdiPW is not shown; composite data-eyes shown would violate TdiPW.
VCENTDQ,midpoint is not shown but is assumed to be midpoint of VdiVW.
The previous figure shows the basic Rx mask requirements. Converting the Rx mask re-
quirements to a classical DQ-to-DQS relationship is shown in the following figure. It
should become apparent that DRAM write training is required to take full advantage of
the Rx mask.
WRITE Operation
Figure 166: Rx Mask DQ-to-DQS DRAM-Based Timings
DQS, DQs Data-In at DRAM Ball DQS, DQs Data-In at DRAM Ball
Rx Mask vs. Composite Data-Eye Rx Mask vs. UI Data-Eye
DQS_c DQS_c
DQS_t DQS_t
TdiPW tDSx tDHx
DRAMa Rx Mask DRAMa Rx Mask
VdiVW
VdiVW
DQx , y, z TdiVW DQx–z TdiVW
TdiPW TdiPW
tDSy tDHy
*Skew
DRAMb Rx Mask
VdiVW
tDQ2DQ
DQy TdiVW
DRAMb
VdiVW
tDQ2DQ Rx Mask
DQz TdiVW
TdiPW
tDSz tDHz
*Skew
DRAMc Rx Mask
VdiVW
tDQ2DQ
DQy TdiVW
DRAMc
VdiVW
Rx Mask tDQ2DQ
DQz TdiVW
TdiPW
Notes: 1. DQx represents an optimally centered mask.

DQy represents earliest valid mask.
DQz represents latest valid mask.
2. *Skew = tDQS2DQ + 0.5 × TdiVW
DRAMa represents a DRAM without any DQS/DQ skews.
DRAMb represents a DRAM with the earliest skews (negative tDQS2DQ, tDQSy > *Skew).
DRAMc represents a DRAM with the latest skews (positive tDQS2DQ, tDQHz > *Skew).
3. DS/tDH are traditional data-eye setup/hold edges at DC levels.
t
tDS and tDH are not specified; tDH and tDS may be any value provided the pulse width
and Rx mask limits are not violated.

tDH (MIN) > TdiVW + tDS (MIN) + tDQ2DQ.
The DDR4 SDRAM's input receivers are expected to capture the input data with an Rx
mask of TdiVW provided the minimum pulse width is satisfied. The DRAM controller
will have to train the data input buffer to utilize the Rx mask specifications to this maxi-
WRITE Operation
mum benefit. If the DRAM controller does not train the data input buffers, then the
worst case limits have to be used for the Rx mask (TdiVW + 2 × tDQS2DQ), which will
generally be the classical minimum ( tDS and tDH) and is required as well.
Figure 167: Example of Data Input Requirements Without Training
TdiVW + 2 × tDQS2DQ
VIH(DC)
0.5 × VdiVW
VdiVW
Rx Mask VCENTDQ,midpoint
0.5 × VdiVW
VIL(DC)
tDS tDH
0.5 × TdiVW + tDQS2DQ 0.5 × TdiVW + tDQS2DQ
DQS_c
DQS_t
WRITE Burst Operation

The following write timing diagrams are intended to help understand each write pa-
rameter's meaning and are only examples. Each parameter will be defined in detail sep-
arately. In these write timing diagrams, CK and DQS are shown aligned, and DQS and
DQ are shown center-aligned for the purpose of illustration.
DDR4 WRITE command supports bursts of BL8 (fixed), BC4 (fixed), and BL8/BC4 on-
the-fly (OTF); OTF uses address A12 to control OTF when OTF is enabled:
• A12 = 0, BC4 (BC4 = burst chop)
• A12 = 1, BL8
WRITE commands can issue precharge automatically with a WRITE with auto pre-
charge (WRA) command, which is enabled by A10 HIGH.
• WRITE command with A10 = 0 (WR) performs standard write, bank remains active af-
ter WRITE burst
• WRITE command with A10 = 1 (WRA) performs write with auto precharge, bank goes
into precharge after WRITE burst
The DATA MASK (DM) function is supported for the x8 and x16 configurations only (the
DM function is not supported on x4 devices). The DM function shares a common pin
with the DBI_n and TDQS functions. The DM function only applies to WRITE opera-
tions and cannot be enabled at the same time the DBI function is enabled.
• If DM_n is sampled LOW on a given byte lane, the DRAM masks the write data re-
ceived on the DQ inputs.
WRITE Operation
• If DM_n is sampled HIGH on a given byte lane, the DRAM does not mask the data and
writes this data into the DRAM core.
• If CRC write is enabled, then DM enabled (via MRS) will be selected between write
CRC nonpersistent mode (DM disabled) and write CRC persistent mode (DM ena-
bled).
Figure 168: WRITE Burst Operation, WL = 9 (AL = 0, CWL = 9, BL8)

T0 T1 T2 T7 T8 T9 T10 T11 T12 T13 T14 T15 T16
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES
Bank Group BGa

Address
Address Bank
Col n
tWPST
tWPRE
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
WL = AL + CWL = 9
Notes: 1. BL8, WL = 0, AL = 0, CWL = 9, Preamble = 1tCK.

2. DI n = Data-in from column n.
these times.
5. CA parity = Disable, CS to CA Latency = Disable, Read DBI = Disable.
WRITE Operation
Figure 169: WRITE Burst Operation, WL = 19 (AL = 10, CWL = 9, BL8)

T0 T1 T2 T9 T10 T11 T17 T18 T19 T20 T21 T22 T23
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES
Bank Group BGa

Address
Address Bank
Col n
tWPRE tWPST
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
AL = 10 CWL = 9
WL = AL + CWL = 19
Notes: 1. BL8, WL = 19, AL = 10 (CL - 1), CWL = 9, Preamble = 1tCK.

2. DI n = data-in from column n.
these times.
WRITE Operation Followed by Another WRITE Operation
Figure 170: Consecutive WRITE (BL8) with 1tCK Preamble in Different Bank Group
CK_c
CK_t
Command WRITE DES DES DES WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES
tWR
tCCD_S =4 4 Clocks tWTR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
DQ DI DI DI DI DI DI DI DI DI DI DI DI DI DI DI DI
WL = AL + CWL = 9
Notes: 1. BL8, AL = 0, CWL = 9, Preamble = 1tCK.

2. DI n (or b) = data-in from column n (or column b).
WRITE Operation
these times.
WRITE commands at T0 and T4.
5. CA parity = Disable, CS to CA latency = Disable, Write DBI = Disable, Write CRC = Disable.
6. The write recovery time (tWR) and write timing parameter (tWTR) are referenced from
the first rising clock edge after the last write data shown at T17.
Figure 171: Consecutive WRITE (BL8) with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 10
WL = AL + CWL = 10
Notes: 1. BL8, AL = 0, CWL = 9 + 1 = 10 (see Note 7), Preamble = 2tCK.

these times.
value at least 1 clock greater than the lowest CWL setting supported in the applicable
tCK range, which means CWL = 9 is not allowed when operating in 2tCK WRITE pream-
ble mode.
WRITE Operation
Figure 172: Nonconsecutive WRITE (BL8) with 1tCK Preamble in Same or Different Bank Group
CK_c
CK_t
Command WRITE DES DES DES DES WRITE DES DES DES DES DES DES DES DES DES DES DES DES
tWR
tCCD_S/L =5 4 Clocks tWTR
Bank Group BGa BGa

Address or BGb
Address Bank Bank

Col n Col b
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
WL = AL + CWL = 9
Notes: 1. BL8, AL = 0, CWL = 9, Preamble = 1tCK, tCCD_S/L = 5tCK.

these times.
Figure 173: Nonconsecutive WRITE (BL8) with 2tCK Preamble in Same or Different Bank Group
CK_c
CK_t
Command WRITE DES DES WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES DES
tWR
Bank Group BGa BGa

Address or BGb
Address Bank Bank

Col n Col b
tWPRE tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 10
WL = AL + CWL = 10
Notes: 1. BL8, AL = 0, CWL = 9 + 1 = 10 (see Note 8), Preamble = 2tCK, tCCD_S/L = 6tCK.
these times.
WRITE Operation
6. tCCD_S/L = 5 isn’t allowed in 2tCK preamble mode.
ble mode.
Figure 174: WRITE (BC4) OTF to WRITE (BC4) OTF with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPST tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = AL + CWL = 9
Notes: 1. BC4, AL = 0, CWL = 9, Preamble = 1tCK.

these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE commands at T0 and
T4.
WRITE Operation
Figure 175: WRITE (BC4) OTF to WRITE (BC4) OTF with 2tCK Preamble in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 10
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = AL + CWL = 10
Notes: 1. BC4, AL = 0, CWL = 9 + 1 = 10 (see Note 7), Preamble = 2tCK.

these times.
T4.
ble mode.
Figure 176: WRITE (BC4) Fixed to WRITE (BC4) Fixed with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
WL = AL + CWL = 9
n n+1 n+2 n+3 b b+1 b+2 b+3
WL = AL + CWL = 9
Notes: 1. BC4, AL = 0, CWL = 9, Preamble = 1tCK.

WRITE Operation
these times.
Figure 177: WRITE (BL8) to WRITE (BC4) OTF with 1tCK Preamble in Different Bank Group
CK_c
CK_t
t
WR
t
CCD_S = 4 4 Clocks t
WTR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
t
t
WPRE WPST
DQS_t,
DQS_c
WL = AL + CWL = 9
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3
WL = AL + CWL = 9
Notes: 1. BL = 8/BC = 4, AL = 0, CL = 9, Preamble = 1tCK.

these times.
4. BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during WRITE command at T0.
BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE command at T4.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, Write CRC = Disable.
WRITE Operation
Figure 178: WRITE (BC4) OTF to WRITE (BL8) with 1tCK Preamble in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
WL = AL + CWL = 9
n n+1 n+2 n+3 b b+1 b+2 b+3 b+4 b+5 b+6 b+7
WL = AL + CWL = 9
Notes: 1. BL = 8/BC = 4, AL = 0, CL = 9, Preamble = 1tCK.

these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE command at T0.
BL8 setting activated by MR0[1:0] = 01 and A12 = 1 during WRITE command at T4.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, Write CRC = Disable.
WRITE Operation Followed by READ Operation
Figure 179: WRITE (BL8) to READ (BL8) with 1tCK Preamble in Different Bank Group
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES READ DES DES DES DES DES DES DES
4 Clocks tWTR_S =2
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPST tRPRE
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
DQ DI DI DI DI DI DI DI DI DI DI DI DI DI DI DI
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2 b+3 b+4 b+5 b+6
Notes: 1. BL = 8, WL = 9 (CWL = 9, AL = 0), CL = 11, READ preamble = 1tCK, WRITE preamble =

1tCK.
2. DI b = data-in from column b.
these times.
WRITE Operation

WRITE command at T0 and READ command at T15.
6. The write timing parameter (tWTR_S) is referenced from the first rising clock edge after
the last write data shown at T13.
Figure 180: WRITE (BL8) to READ (BL8) with 1tCK Preamble in Same Bank Group
CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES READ DES DES DES DES DES
4 Clocks tWTR_L =4
Bank Group BGa BGa

Address
Address Bank Bank

Col n Col b
tWPRE tWPST tRPRE
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
DQ DI DI DI DI DI DI DI DI DI DI DI
n n+1 n+2 n+3 n+4 n+5 n+6 n+7 b b+1 b+2
Notes: 1. BL = 8, WL = 9 (CWL = 9, AL = 0), CL = 11, READ preamble = 1tCK, WRITE preamble =

1tCK.
these times.
WRITE command at T0 and READ command at T17.
6. The write timing parameter (tWTR_L) is referenced from the first rising clock edge after
WRITE Operation
Figure 181: WRITE (BC4) OTF to READ (BC4) OTF with 1tCK Preamble in Different Bank Group
CK_c
CK_t
4 Clocks tWTR_S =2
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
tWPRE tWPST tRPRE tRPST
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BC = 4, WL = 9 (CWL = 9, AL = 0), CL = 11, READ preamble = 1tCK, WRITE preamble =

1tCK.
these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE command at T0 and
READ command at T15.
Figure 182: WRITE (BC4) OTF to READ (BC4) OTF with 1tCK Preamble in Same Bank Group
CK_c
CK_t
4 Clocks tWTR_L =4
Bank Group BGa BGa

Address
Address Bank Bank

Col n Col b
tWPRE tWPST tRPRE
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
DQ DI DI DI DI DI DI DI
n n+1 n+2 n+3 b b+1 b+2
Notes: 1. BC = 4, WL = 9 (CWL = 9, AL = 0), CL = 11, READ preamble = 1tCK, WRITE preamble =

1tCK.
these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE command at T0 and
READ command at T17.
WRITE Operation
Figure 183: WRITE (BC4) Fixed to READ (BC4) Fixed with 1 tCK Preamble in Different Bank Group
CK_c
CK_t
2 Clocks tWTR_S =2
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BC = 4, WL = 9 (CWL = 9, AL = 0), CL = 11, READ preamble = 1 tCK, WRITE preamble =

1tCK.
these times.
Figure 184: WRITE (BC4) Fixed to READ (BC4) Fixed with 1tCK Preamble in Same Bank Group
CK_c
CK_t
2 Clocks tWTR_L =4
Bank Group BGa BGa

Address
Address Bank Bank

Col n Col b
DQS_t,
DQS_c
WL = AL + CWL = 9 RL = AL + CL = 11
n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BC = 4, WL = 9 (CWL = 9, AL = 0), C L = 11, READ preamble = 1tCK, WRITE preamble =

1tCK.
WRITE Operation
these times.
WRITE Operation Followed by PRECHARGE Operation

The minimum external WRITE command to PRECHARGE command spacing is equal to
WL (AL + CWL) plus either 4tCK (BL8/BC4-OTF) or 2tCK (BC4-fixed) plus tWR. The min-
imum ACT to PRE timing, tRAS, must be satisfied as well.
Figure 185: WRITE (BL8/BC4-OTF) to PRECHARGE with 1tCK Preamble

CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES PRE DES
WL = AL + CWL = 9 4 Clocks tWR = 12 tRP
BGa, Bank b BGa, Bank b

Col n (or all)
Address
BC4 (OTF) Opertaion

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. BL = 8 with BC4-OTF, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK, tWR = 12.

these times.
4. BC4 setting activated by MR0[1:0] = 01 and A12 = 0 during WRITE command at T0. BL8
setting activated by MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during WRITE command
at T0.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, CRC = Disable.
6. The write recovery time (tWR) is referenced from the first rising clock edge after the last
write data shown at T13. tWR specifies the last burst WRITE cycle until the PRECHARGE
command can be issued to the same bank.
WRITE Operation
Figure 186: WRITE (BC4-Fixed) to PRECHARGE with 1tCK Preamble

CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES PRE DES DES DES
BGa, Bank b BGa, Bank b

Col n (or all)
Address
BC4 (Fixed) Opertaion

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
Notes: 1. BC4 = fixed, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK, tWR = 12.

these times.
Figure 187: WRITE (BL8/BC4-OTF) to Auto PRECHARGE with 1tCK Preamble

CK_c
CK_t
Command WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES
BGa, Bank b
Col n
Address
BC4 (OTF) Opertaion

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. BL = 8 with BC4-OTF, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK, tWR = 12.

these times.
BL8 setting activated by MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during WRITE com-
mand at T0.
WRITE Operation
Figure 188: WRITE (BC4-Fixed) to Auto PRECHARGE with 1tCK Preamble

CK_c
CK_t
BGa, Bank b
Col n
Address

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
Notes: 1. BC4 = fixed, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK, tWR = 12.

these times.
WRITE Operation
WRITE Operation with WRITE DBI Enabled
Figure 189: WRITE (BL8/BC4-OTF) with 1tCK Preamble and DBI

CK_c
CK_t
WL = AL + CWL = 9 4 Clocks tWR
tWTR
Address BGa
Address Bank,
Col n
BC4 (OTF) Opertaion

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
DBI_n DI DI DI DI
n n+1 n+2 n+3
BL8 Opertaion
DQS_t,
DQS_c
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
DBI_n DI DI DI DI DI DI DI DI
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
Notes: 1. BL = 8 with BC4-OTF, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK.

these times.
BL8 setting activated by MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during WRITE com-
mand at T0.
5. CA parity = Disable, CS to CA latency = Disable, Write DBI = Enabled, Write CRC = Disa-
bled.
6. The write recovery time (tWR_DBI) is referenced from the first rising clock edge after the
last write data shown at T13.
WRITE Operation
Figure 190: WRITE (BC4-Fixed) with 1tCK Preamble and DBI

CK_c
CK_t
WL = AL + CWL = 9 2 Clocks tWR
tWTR
Address BGa
Address Bank,
Col n

DQS_t,
DQS_c
DQ DI DI DI DI
n n+1 n+2 n+3
DBI_n DI DI DI DI
n n+1 n+2 n+3
Notes: 1. BC4 = fixed, WL = 9 (CWL = 9, AL = 0 ), Preamble = 1tCK.

these times.
5. CA parity = Disable, CS to CA latency = Disable, Write DBI = Enabled, Write CRC = Disa-
bled.
WRITE Operation
WRITE Operation with CA Parity Enabled
Figure 191: Consecutive Write (BL8) with 1tCK Preamble and CA Parity in Different Bank Group
CK_c
CK_t
tWR
Bank Group BGa BGb

Address
Address Bank Bank

Col n Col b
Parity Valid Valid

tWPRE tWPST
DQS_t,
DQS_c
WL = PL + AL + CWL = 13
WL = PL + AL + CWL = 13
Notes: 1. BL = 8, WL = 9 (CWL = 13, AL = 0 ), Preamble = 1tCK.

these times.
4. BL8 setting activated by MR0[1:0] = 00 or MR0[1:0] = 01 and A12 = 1 during WRITE com-
mands at T0 and T4.
5. CA parity = Enable, CS to CA latency = Disable, Write DBI = Enabled, Write CRC = Disa-
ble.
WRITE Operation
WRITE Operation with Write CRC Enabled
Figure 192: Consecutive WRITE (BL8/BC4-OTF) with 1tCK Preamble and Write CRC in Same or Differ-
ent Bank Group
CK_c
CK_t
tWR

Address BGb
Address Bank Bank

Col n Col b
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
DQ x4, DI DI DI DI DI DI DI DI CRC CRC DI DI DI DI DI DI DI DI CRC CRC

BL = 8
WL = AL + CWL = 9
DQ x8/X16, DI DI DI DI DI DI DI DI CRC DI DI DI DI DI DI DI DI CRC

BL = 8
DQ x4, DI DI DI DI DI DI DI DI
BC = 4 (OTF) n n+1 n+2 n+3 CRC CRC b b+1 b+2 b+3 CRC CRC
DQ x8/X16,
DI DI DI DI CRC DI DI DI DI CRC
BC = 4 (OTF) n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BL8/BC4-OTF, AL = 0, CWL = 9, Preamble = 1tCK, tCCD_S/L = 5tCK.

these times.
T5.
6. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, Write CRC = Enable.
WRITE Operation
Figure 193: Consecutive WRITE (BC4-Fixed) with 1tCK Preamble and Write CRC in Same or Different
Bank Group
CK_c
CK_t
tWR

Address BGb
Address Bank Bank

Col n Col b
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
BC = 4 (Fixed) n n+1 n+2 n+3 CRC CRC b b+1 b+2 b+3 CRC CRC
WL = AL + CWL = 9
DQ x8/X16,
BC = 4 (Fixed) n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BC4-fixed, AL = 0, CWL = 9, Preamble = 1tCK, tCCD_S/L = 5tCK.

these times.
4. BC4 setting activated by MR0[1:0] = 10 during WRITE commands at T0 and T5.
5. CA parity = Disable, CS to CA latency = Disable, Read DBI = Disable, Write CRC = Enable,
DM = Disable.
WRITE Operation
Figure 194: Nonconsecutive WRITE (BL8/BC4-OTF) with 1tCK Preamble and Write CRC in Same or Dif-
ferent Bank Group
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
&.BF
&.BW
&RPPDQG :5,7( '(6 '(6 :5,7( '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6 '(6
W :5
W &&'B6/ &ORFNV W :75
%DQN*URXS %*D %*DRU

$GGUHVV %*E
$GGUHVV %DQN %DQN

&ROQ &ROE
W :35( W :367
'46BW
'46BF
:/ $/&:/
'4[ ', ', ', ', ', ', ', ', &5& &5& ', ', ', ', ', ', ', ', &5& &5&
Q Q Q Q Q Q Q Q E E E E E E E E
%/
:/ $/&:/
'4[; ', ', ', ', ', ', ', ', &5& ', ', ', ', ', ', ', ', &5&
Q Q Q Q Q Q Q Q E E E E E E E E
%/
'4[ ', ', ', ', ', ', ', ',

%& 27) Q Q Q Q &5& &5& E E E E &5& &5&
'4[;
', ', ', ', &5& ', ', ', ', &5&
%& 27) Q Q Q Q E E E E
7LPH%UHDN 7UDQVLWLRQLQJ'DWD 'RQ¶W&DUH
Notes: 1. BL8/BC4-OTF, AL = 0, CWL = 9, Preamble = 1tCK, tCCD_S/L = 6tCK.

these times.
T6.
DM = Disable.
WRITE Operation
Figure 195: Nonconsecutive WRITE (BL8/BC4-OTF) with 2tCK Preamble and Write CRC in Same or Dif-
ferent Bank Group
CK_c
CK_t
Command WRITE DES WRITE DES DES DES DES DES DES DES DES DES DES DES DES DES DES DES
tWR

Address BGb
Address Bank Bank

Col n Col b
tWPRE tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 10
DQ x4, DI DI DI DI DI DI DI DI CRC CRC DI DI DI DI DI DI DI DI CRC CRC

BL = 8
WL = AL + CWL = 10
DQ x8/X16, DI DI DI DI DI DI DI DI CRC DI DI DI DI DI DI DI DI CRC

BL = 8
BC = 4 (OTF) n n+1 n+2 n+3 CRC CRC b b+1 b+2 b+3 CRC CRC
DQ x8/X16,
BC = 4 (OTF) n n+1 n+2 n+3 b b+1 b+2 b+3
Notes: 1. BL8/BC4-OTF, AL = 0, CWL = 9 + 1 = 10 (see Note 9), Preamble = 2tCK, tCCD_S/L = 7tCK
(see Note 7).
these times.
T7.
DM = Disable.
7. tCCD_S/L = 6tCK is not allowed in 2tCK preamble mode if minimum tCCD_S/L allowed in
1tCK preamble mode would have been 6 clocks.
tCK range. That means CWL = 9 is not allowed when operating in 2tCK WRITE preamble
mode.
WRITE Operation
Figure 196: WRITE (BL8/BC4-OTF/Fixed) with 1tCK Preamble and Write CRC in Same or Different Bank
Group
CK_c
CK_t
tWR_CRC_DM
4 Clocks tWTR_S_CRC_DM/tWTR_L_CRC_DM
Bank Group BGa

Address
Address Bank
Col n
tWPRE tWPST
DQS_t,
DQS_c
WL = AL + CWL = 9
DQ x4, DI DI DI DI DI DI DI DI CRC CRC

n n+1 n+2 n+3 n+4 n+5 n+6 n+7
BL = 8
DQ x8/X16, DI DI DI DI DI DI DI DI CRC
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
BL = 8
DMx4/x8/x16 DM DM DM DM DM DM DM DM
n n+1 n+2 n+3 n+4 n+5 n+6 n+7
BL = 8
DQ x4, DI DI DI DI
BC = 4 (OTF/Fixed) n n+1 n+2 n+3 CRC CRC
DQ x8/X16,
DI DI DI DI CRC
BC = 4 (OTF/Fixed) n n+1 n+2 n+3
DM x4/x8/x16
DM DM DM DM
BC = 4 (OTF / Fixed) n n+1 n+2 n+3
Notes: 1. BL8/BC4, AL = 0, CWL = 9, Preamble = 1tCK.

these times.
5. BC4 setting activated by either MR0[1:0] = 10 or MR0[1:0] = 01 and A12 = 0 during
DM = Enable.
7. The write recovery time (tWR_CRC_DM) and write timing parameter (tWTR_S_CRC_DM/
tWTR_L_CRC_DM) are referenced from the first rising clock edge after the last write da-
ta shown at T13.
Write Timing Violations
Write Timing Violations

Motivation
Generally, if timing parameters are violated, a complete reset/initialization procedure
has to be initiated to make sure that the device works properly. However, for certain mi-
nor violations, it is desirable that the device is guaranteed not to "hang up" and that er-
rors are limited to that specific operation. A minor violation does not include a major
timing violation (for example, when a DQS strobe misses in the tDQSCK window).
For the following, it will be assumed that there are no timing violations with regard to
the WRITE command itself (including ODT, and so on) and that it does satisfy all timing
requirements not mentioned below.
Data Setup and Hold Violations

If the data-to-strobe timing requirements (tDS, tDH) are violated, for any of the strobe
edges associated with a WRITE burst, then wrong data might be written to the memory
location addressed with this WRITE command.
In the example, the relevant strobe edges for WRITE Burst A are associated with the
clock edges: T5, T5.5, T6, T6.5, T7, T7.5, T8, and T8.5.
Subsequent reads from that location might result in unpredictable read data; however,
the device will work properly otherwise.
Strobe-to-Strobe and Strobe-to-Clock Violations

If the strobe timing requirements (tDQSH, tDQSL, tWPRE, tWPST) or the strobe to clock
timing requirements (tDSS, tDSH, tDQSS) are violated, for any of the strobe edges asso-
ciated with a WRITE burst, then wrong data might be written to the memory location
addressed with the offending WRITE command. Subsequent reads from that location
might result in unpredictable read data; however, the device will work properly other-
wise with the following constraints:
• Both write CRC and data burst OTF are disabled; timing specifications other than
tDQSH, tDQSL, tWPRE, tWPST, tDSS, tDSH, tDQSS are not violated.
• The offending write strobe (and preamble) arrive no earlier or later than six DQS tran-
sition edges from the WRITE latency position.
• A READ command following an offending WRITE command from any open bank is
allowed.
• One or more subsequent WR or a subsequent WRA (to same bank as offending WR)
may be issued tCCD_L later, but incorrect data could be written. Subsequent WR and
WRA can be either offending or non-offending writes. Reads from these writes may
provide incorrect data.
• One or more subsequent WR or a subsequent WRA (to a different bank group) may be
issued tCCD_S later, but incorrect data could be written. Subsequent WR and WRA
can be either offending or non-offending writes. Reads from these writes may provide
incorrect data.
• After one or more precharge commands (PRE or PREA) are issued to the device after
an offending WRITE command and all banks are in precharged state (idle state), a
subsequent, non-offending WR or WRA to any open bank will be able to write correct
data.
ZQ CALIBRATION Commands
A ZQ CALIBRATION command is used to calibrate DRAM RON and ODT values. The de-
vice needs a longer time to calibrate the output driver and on-die termination circuits
at initialization and a relatively smaller time to perform periodic calibrations.
The ZQCL command is used to perform the initial calibration during the power-up ini-
tialization sequence. This command may be issued at any time by the controller de-
pending on the system environment. The ZQCL command triggers the calibration en-
gine inside the DRAM and, after calibration is achieved, the calibrated values are trans-
ferred from the calibration engine to DRAM I/O, which is reflected as an updated out-
put driver and ODT values.
The first ZQCL command issued after reset is allowed a timing period of tZQinit to per-
form the full calibration and the transfer of values. All other ZQCL commands except
the first ZQCL command issued after reset are allowed a timing period of tZQoper.
The ZQCS command is used to perform periodic calibrations to account for voltage and
temperature variations. A shorter timing window is provided to perform the calibration
and transfer of values as defined by timing parameter tZQCS. One ZQCS command can
effectively correct a minimum of 0.5% (ZQ correction) of RON and RTT impedance error
within 64 nCK for all speed bins assuming the maximum sensitivities specified in the
Output Driver and ODT Voltage and Temperature Sensitivity tables. The appropriate in-
terval between ZQCS commands can be determined from these tables and other appli-
cation-specific parameters. One method for calculating the interval between ZQCS
commands, given the temperature (Tdrift_rate) and voltage (Vdrift_rate) drift rates that the
device is subjected to in the application, is illustrated. The interval could be defined by
the following formula:
ZQcorrection
(Tsense x Tdrift_rate) + (Vsense x Tdrift_rate)
Where T sense = MAX(dRTTdT, dRONdTM) and V sense = MAX(dRTTdV, dRONdVM) define

the temperature and voltage sensitivities.
For example, if T sens = 1.5%/°C, V sens = 0.15%/mV, T driftrate = 1 °C/sec and V driftrate = 15
mV/sec, then the interval between ZQCS commands is calculated as:
0.5
= 0.133 §128ms
(1.5 × 1) + (0.15 × 15)
No other activities should be performed on the DRAM channel by the controller for the
duration of tZQinit, tZQoper, or tZQCS. The quiet time on the DRAM channel allows ac-
curate calibration of output driver and on-die termination values. After DRAM calibra-
tion is achieved, the device should disable the ZQ current consumption path to reduce
power.
All banks must be precharged and tRP met before ZQCL or ZQCS commands are issued
by the controller.
ZQ CALIBRATION commands can also be issued in parallel to DLL lock time when
coming out of self refresh. Upon self refresh exit, the device will not perform an I/O cali-
bration without an explicit ZQ CALIBRATION command. The earliest possible time for a
ZQ CALIBRATION command (short or long) after self refresh exit is tXS, tXS_Abort, or
tXS_FAST depending on operation mode.
In systems that share the ZQ resistor between devices, the controller must not allow any
overlap of tZQoper, tZQinit, or tZQCS between the devices.
Figure 197: ZQ Calibration Timing
T0 T1 Ta0 Ta1 Ta2 Ta3 Tb0 Tb1 Tc0 Tc1 Tc2

CK_c
CK_t
Command ZQCL DES DES DES Valid Valid ZQCS DES DES DES Valid
A10 Valid Valid Valid

Note 1
Note 2
ODT Valid Valid Valid
DQ Bus High-Z or RTT(Park) Activities High-Z or RTT(Park) Activities

Note 3
tZQinit_tZQoper tZQCS
Notes: 1. CKE must be continuously registered HIGH during the calibration procedure.
2. During ZQ calibration, the ODT signal must be held LOW and DRAM continues to pro-
vide RTT_PARK.
3. All devices connected to the DQ bus should be High-Z during the calibration procedure.
On-Die Termination
On-Die Termination
The on-die termination (ODT) feature enables the device to change termination resist-
ance for each DQ, DQS, and DM_n/DBI_n signal for x4 and x8 configurations (and
TDQS for the x8 configuration when enabled via A11 = 1 in MR1) via the ODT control
pin, WRITE command, or default parking value with MR setting. For the x16 configura-
tion, ODT is applied to each UDQ, LDQ, UDQS, LDQS, UDM_n/UDBI_n, and LDM_n/
LDBI_n signal. The ODT feature is designed to improve the signal integrity of the mem-
ory channel by allowing the DRAM controller to independently change termination re-
sistance for any or all DRAM devices. If DBI read mode is enabled while the DRAM is in
standby, either DM mode or DBI write mode must also be enabled if RTT(NOM) or
RTT(Park) is desired. More details about ODT control modes and ODT timing modes can
be found further along in this document.
The ODT feature is turned off and not supported in self refresh mode.
Figure 198: Functional Representation of ODT
ODT VDDQ
To other
circuitry RTT
such as
RCV, Switch
... DQ, DQS, DM, TDQS
The switch is enabled by the internal ODT control logic, which uses the external ODT
pin and other control information. The value of R TT is determined by the settings of
mode register bits (see Mode Register). The ODT pin will be ignored if the mode register
MR1 is programmed to disable RTT(NOM) [MR1[10,9,8] = 0,0,0] and in self refresh mode.
ODT Mode Register and ODT State Table

The ODT mode of the DDR4 device has four states: data termination disable, RTT(NOM),
RTT(WR), and RTT(Park). The ODT mode is enabled if any of MR1[10:8] (R TT(NOM)),
MR2[11:9] (RTT(WR)), or MR5[8:6] (RTT(Park)) are non-zero. When enabled, the value of
RTT is determined by the settings of these bits.
RTT control of each RTT condition is possible with a WR or RD command and ODT pin.
• RTT(WR): The DRAM (rank) that is being written to provide termination regardless of
ODT pin status (either HIGH or LOW).
• RTT(NOM): DRAM turns ON RTT(NOM) if it sees ODT asserted HIGH (except when ODT
is disabled by MR1).
• RTT(Park): Default parked value set via MR5 to be enabled and RTT(NOM) is not turned
on.
• The Termination State Table that follows shows various interactions.
The RTT values have the following priority:
• Data termination disable
• RTT(WR)
• RTT(NOM)
• RTT(Park)
ODT Mode Register and ODT State Table
Table 70: Termination State Table
Case RTT(Park) RTT(NOM)1 RTT(WR)2 ODT Pin ODT READS3 ODT Standby ODT WRITES
A4 Disabled Disabled Disabled Don't Care Off (High-Z) Off (High-Z) Off (High-Z)
Enabled Don't Care Off (High-Z) Off (High-Z) RTT(WR)
B5 Enabled Disabled Disabled Don't Care Off (High-Z) RTT(Park) RTT(Park)
Enabled Don't Care Off (High-Z) RTT(Park) RTT(WR)
C6 Disabled Enabled Disabled Low Off (High-Z) Off (High-Z) Off (High-Z)
High Off (High-Z) RTT(NOM) RTT(NOM)
Enabled Low Off (High-Z) Off (High-Z) RTT(WR)
High Off (High-Z) RTT(NOM) RTT(WR)
D6 Enabled Enabled Disabled Low Off (High-Z) RTT(Park) RTT(Park)
High Off (High-Z) RTT(NOM) RTT(NOM)
Enabled Low Off (High-Z) RTT(Park) RTT(WR)
High Off (High-Z) RTT(NOM) RTT(WR)
Notes: 1. If RTT(NOM) MR is disabled, power to the ODT receiver will be turned off to save power.
2. If RTT(WR) is enabled, RTT(WR) will be activated by a WRITE command for a defined period
time independent of the ODT pin and MR setting of RTT(Park)/RTT(NOM). This is described in
the Dynamic ODT section.
3. When a READ command is executed, the DRAM termination state will be High-Z for a
defined period independent of the ODT pin and MR setting of RTT(Park)/RTT(NOM). This is
described in the ODT During Read section.
4. Case A is generally best for single-rank memories.
5. Case B is generally best for dual-rank, single-slotted memories.
6. Case C and Case D are generally best for multi-slotted memories.
ODT Read Disable State Table

Upon receiving a READ command, the DRAM driving data disables ODT after RL - (2 or
3) clock cycles, where 2 = 1tCK preamble mode and 3 = 2tCK preamble mode. ODT stays
off for a duration of BL/2 + (2 or 3) + (0 or 1) clock cycles, where 2 = 1tCK preamble
mode, 3 = 2tCK preamble mode, 0 = CRC disabled, and 1 = CRC enabled.
Table 71: Read Termination Disable Window
Start ODT Disable After

Preamble CRC Read Duration of ODT Disable
1tCK Disabled RL - 2 BL/2 + 2
Enabled RL - 2 BL/2 + 3
2tCK Disabled RL - 3 BL/2 + 3
Enabled RL - 3 BL/2 + 4
Synchronous ODT Mode

Synchronous ODT mode is selected whenever the DLL is turned on and locked. Based
on the power-down definition, these modes include the following:
• Any bank active with CKE HIGH
• Refresh with CKE HIGH
• Idle mode with CKE HIGH
• Active power-down mode (regardless of MR1 bit A10)
• Precharge power-down mode
In synchronous ODT mode, RTT(NOM) will be turned on DODTLon clock cycles after
ODT is sampled HIGH by a rising clock edge and turned off DODTLoff clock cycles after
ODT is registered LOW by a rising clock edge. The ODT latency is determined by the
programmed values for: CAS WRITE latency (CWL), additive latency (AL), and parity la-
tency (PL), as well as the programmed state of the preamble.
ODT Latency and Posted ODT

The ODT latencies for synchronous ODT mode are summarized in the table below. For
details, refer to the latency definitions.
Table 72: ODT Latency at DDR4-1600/-1866/-2133/-2400/-2666/-3200

Applicable when write CRC is disabled
Symbol Parameter 1tCK Preamble 2tCK Preamble Unit
DODTLon Direct ODT turn-on latency CWL + AL + PL - 2 CWL + AL + PL - 3 tCK
DODTLoff Direct ODT turn-off latency CWL + AL + PL - 2 CWL + AL + PL - 3

RODTLoff READ command to internal ODT turn-off CL + AL + PL - 2 CL + AL + PL - 3
latency
RODTLon4 READ command to RTT(Park) turn-on la- RODTLoff + 4 RODTLoff + 5
tency in BC4-fixed
RODTLon8 READ command to RTT(Park) turn-on la- RODTLoff + 6 RODTLoff + 7
tency in BL8/BC4-OTF
ODTH4 ODT Assertion time, BC4 mode 4 5
ODTH8 ODT Assertion time, BL8 mode 6 7
Timing Parameters
In synchronous ODT mode, the following parameters apply:
• DODTLon, DODTLoff, RODTLoff, RODTLon4, RODTLon8, and tADC (MIN)/(MAX).
• tADC (MIN) and tADC (MAX) are minimum and maximum RTT change timing skew
between different termination values. These timing parameters apply to both the syn-
chronous ODT mode and the data termination disable mode.
When ODT is asserted, it must remain HIGH until minimum ODTH4 (BC = 4) or
ODTH8 (BL = 8) is satisfied. If write CRC mode or 2tCK preamble mode is enabled,
ODTH should be adjusted to account for it. ODTHx is measured from ODT first regis-
tered HIGH to ODT first registered LOW or from the registration of a WRITE command.
Figure 199: Synchronous ODT Timing with BL8
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
diff_CK
Command
ODT
DODTLon = WL - 2 DODTLoff = WL - 2
t ADC (MAX) tADC (MAX)

tADC (MIN) tADC (MIN)
DRAM_RTT RTT(Park) RTT(NOM) RTT(Park)
Transitioning
Notes: 1. Example for CWL = 9, AL = 0, PL = 0; DODTLon = AL + PL + CWL - 2 = 7; DODTLoff = AL +

PL + CWL - 2 = 7.
2. ODT must be held HIGH for at least ODTH8 after assertion (T1).
Figure 200: Synchronous ODT with BC4
T0 T1 T2 T3 T4 T5 T18 T19 T20 T21 T22 T23 T36 T37 T38 T39 T40 T41 42
diff_CK
WRS4
Command
ODTH4
ODT
DODTLoff = WL - 2
ODTLcnw= WL - 2
DODTLon = CWL - 2 ODTLcwn4 = ODTLcnw + 4
tADC (MAX) tADC (MAX) tADC (MAX) tADC (MAX)

tADC (MIN) tADC (MIN) tADC (MIN) tADC (MIN)
RTT(Park) RTT(NOM) RTT(Park) RTT(WR) RTT(Park)

DRAM_RTT
Transitioning
Notes: 1. Example for CWL = 9, AL = 10, PL = 0; DODTLon/off = AL + PL+ CWL - 2 = 17; ODTcnw =
AL + PL+ CWL - 2 = 17.
2. ODT must be held HIGH for at least ODTH4 after assertion (T1).
ODT During Reads

Because the DRAM cannot terminate with RTT and drive with RON at the same time, RTT
may nominally not be enabled until the end of the postamble as shown in the example
below. At cycle T26 the device turns on the termination when it stops driving, which is
determined by tHZ. If the DRAM stops driving early (that is, tHZ is early), then tADC
(MIN) timing may apply. If the DRAM stops driving late (that is, tHZ is late), then the
DRAM complies with tADC (MAX) timing.
Using CL = 11 as an example for the figure below: PL = 0, AL = CL - 1 = 10, RL = PL + AL +
CL = 21, CWL= 9; RODTLoff = RL - 2 = 19, DODTLon = PL + AL + CWL - 2 = 17, 1tCK
preamble.
Figure 201: ODT During Reads
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
GLIIB&.
&RPPDQG 5'
$GGUHVV $
5/ $/&/3/
2'7
52'7/RII 5/
'2'7/RQ :/
W $'&0$; W $'&0$;
W $'&0,1 W $'&0,1
'46B2'7 5773DUN 577120

W &.3UHDPEOH
W $'&0$; W $'&0$;
W $'&0,1 W $'&0,1
'46B2'7 5773DUN 577120

W &.3UHDPEOH
'46GLII
W $'&0$;Q&. W $'&0$;
W $'&0,1 W $'&0,1
'4B2'7 5773DUN 577120

W &.3UHDPEOH
'4 4$ 4$ 4$ 4$ 4$ 4$ 4$ 4$
W $'&0$;Q&. W $'&0$;
W $'&0,1 W $'&0,1
'4B2'7 5773DUN 577120

W &.3UHDPEOH
'4 4$ 4$ 4$ 4$ 4$ 4$ 4$ 4$
7UDQVLWLRQLQJ
Dynamic ODT
Dynamic ODT
In certain application cases and to further enhance signal integrity on the data bus, it is
desirable that the termination strength of the device can be changed without issuing an
MRS command. This requirement is supported by the dynamic ODT feature.
Dynamic ODT mode is enabled if bit A9 or A10 of MR2 is set to 1.
• Three RTT values are available: RTT(NOM), RTT(WR), and RTT(Park).
– The value for RTT(NOM) is preselected via bits MR1[10:8].
– The value for RTT(WR) is preselected via bits MR2[11:9].
– The value for RTT(Park) is preselected via bits MR5[8:6].
• During operation without WRITE commands, the termination is controlled as fol-
lows:
– Nominal termination strength RTT(NOM) or RTT(Park) is selected.
– RTT(NOM) on/off timing is controlled via ODT pin and latencies DODTLon and
DODTLoff, and RTT(Park) is on when ODT is LOW.
• When a WRITE command (WR, WRA, WRS4, WRS8, WRAS4, and WRAS8) is regis-
tered, and if dynamic ODT is enabled, the termination is controlled as follows:
– Latency ODTLcnw after the WRITE command, termination strength R TT(WR) is se-
lected.
– Latency ODTLcwn8 (for BL8, fixed by MRS or selected OTF) or ODTLcwn4 (for
BC4, fixed by MRS or selected OTF) after the WRITE command, termination
strength RTT(WR) is de-selected.
One or two clocks will be added into or subtracted from ODTLcwn8 and ODTLcwn4,
depending on write CRC mode and/or 2tCK preamble enablement.
The following table shows latencies and timing parameters relevant to the on-die termi-
nation control in dynamic ODT mode. The dynamic ODT feature is not supported in
DLL-off mode. An MRS command must be used to set RTT(WR) to disable dynamic ODT
externally (MR2[11:9] = 000).
Table 73: Dynamic ODT Latencies and Timing (1tCK Preamble Mode and CRC Disabled)
Name and Descrip- 1600/1866/

tion Abbr. Defined from Defined to 2133/2400 2666 2933/3200 Unit
ODT latency for ODTLc Registering ex- Change RTT ODTLcnw = WL - 2 tCK
change from RTT(Park)/ nw ternal WRITE strength from

RTT(NOM) to RTT(WR) command RTT(Park)/
RTT(NOM) to
RTT(WR)
ODT latency for ODTLc Registering ex- Change RTT ODTLcwn4 = 4 + ODTLcnw tCK
change from RTT(WR) to wn4 ternal WRITE strength from

RTT(Park)/RTT(NOM) (BC = command RTT(WR) to
4) RTT(Park)/
RTT(NOM)
ODT latency for ODTLc Registering ex- Change RTT ODTLcwn8 = 6 + ODTLcnw tCK
change from RTT(WR) to wn8 ternal WRITE strength from (AVG)

RTT(Park)/RTT(NOM) (BL = command RTT(NOM) to
8) RTT(WR)
Dynamic ODT
Table 73: Dynamic ODT Latencies and Timing (1tCK Preamble Mode and CRC Disabled) (Continued)
Name and Descrip- 1600/1866/

tion Abbr. Defined from Defined to 2133/2400 2666 2933/3200 Unit
RTT change skew tADC ODTLcnw RTT valid tADC (MIN) = tADC (MIN) = tADC (MIN) = tCK
ODTLcwn 0.30 0.28 0.26 (AVG)

tADC (MAX) = tADC (MAX) = tADC (MAX) =
0.70 0.72 0.74
Table 74: Dynamic ODT Latencies and Timing with Preamble Mode and CRC Mode Matrix
1tCK Parameter 2tCK Parameter

Symbol CRC Off CRC On CRC Off CRC On Unit
ODTLcnw1 WL - 2 WL - 2 WL - 3 WL - 3 tCK
ODTLcwn4 ODTLcnw + 4 ODTLcnw + 7 ODTLcnw + 5 ODTLcnw + 8

ODTLcwn8 ODTLcnw + 6 ODTLcnw + 7 ODTLcnw + 7 ODTLcnw + 8
Note: 1. ODTLcnw = WL - 2 (1tCK preamble) or WL - 3 (2tCK preamble).
Figure 202: Dynamic ODT (1t CK Preamble; CL = 14, CWL = 11, BL = 8, AL = 0, CRC Disabled)
T0 T1 T2 T5 T6 T7 T8 T9 T10 T11 T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24
diff_CK
WR
Command
ODT
DODTLon = WL - 2
DODTLoff = WL - 2
tADC (MAX) tADC (MAX) tADC (MAX) tADC (MAX)
RTT RTT(Park) RTT(WR) RTT(Park) RTT(NOM) RTT(Park)
tADC (MIN) tADC (MIN) tADC (MIN) tADC (MIN)

ODTLcnw
ODTLcwn
Transitioning
Notes: 1. ODTLcnw = WL - 2 (1tCK preamble) or WL - 3 (2tCK preamble).

2. If BC4, then ODTLcwn = WL + 4 if CRC disabled or WL + 5 if CRC enabled; If BL8, then
ODTLcwn = WL + 6 if CRC disabled or WL + 7 if CRC enabled.
Dynamic ODT
Figure 203: Dynamic ODT Overlapped with RTT(NOM) (CL = 14, CWL = 11, BL = 8, AL = 0, CRC Disabled)
T0 T1 T2 T5 T6 T7 T9 T10 T11 T12 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25
diff_CK
WR
Command
ODT
ODTLcnw
ODTLcwn8
tADC (MAX) tADC (MAX) tADC (MAX)
RTT RTT_NOM RTT_WR RTT_NOM RTT_PARK
tADC (MIN) tADC (MIN) tADC (MIN)
DODTLoff = CWL -2
Note: 1. Behavior with WR command issued while ODT is registered HIGH.
Asynchronous ODT Mode
Asynchronous ODT Mode

Asynchronous ODT mode is selected when the DRAM runs in DLL-off mode. In asyn-
chronous ODT timing mode, the internal ODT command is not delayed by either addi-
tive latency (AL) or the parity latency (PL) relative to the external ODT signal (RTT(NOM)).
In asynchronous ODT mode, two timing parameters apply: tAONAS (MIN/MAX), and
tAOFAS (MIN/MAX).
RTT(NOM) Turn-on Time

• Minimum RTT(NOM) turn-on time (tAONAS [MIN]) is when the device termination cir-
cuit leaves RTT(Park) and ODT resistance begins to turn on.
• Maximum RTT(NOM) turn-on time (tAONAS [MAX]) is when the ODT resistance has
reached RTT(NOM).
• tAONAS (MIN) and tAONAS (MAX) are measured from ODT being sampled HIGH.
RTT(NOM) Turn-off Time
• Minimum RTT(NOM) turn-off time (tAOFAS [MIN]) is when the device's termination
circuit starts to leave RTT(NOM).
• Maximum RTT(NOM) turn-off time (tAOFAS [MAX]) is when the on-die termination has
reached RTT(Park).
• tAOFAS (MIN) and tAOFAS (MAX) are measured from ODT being sampled LOW.
Figure 204: Asynchronous ODT Timings with DLL Off

T0 T1 T2 T3 T4 T5 T6 Ti Ti + 1 Ti + 2 Ti + 3 Ti + 4 Ti + 5 Ti + 6 Ta Tb
diff_CK
CKE
tIH tIS tIH tIS
ODT
tAONAS (MAX)
tAONAS (MIN)
RTT RTT(Park) RTT(NOM)
tAONAS (MIN)
tAONAS (MAX)
Transitioning
Electrical Specifications
Electrical Specifications
Absolute Ratings
Stresses greater than those listed may cause permanent damage to the device. This is a
stress rating only, and functional operation of the device at these or any other condi-
tions outside those indicated in the operational sections of this specification is not im-
plied. Exposure to absolute maximum rating conditions for extended periods may ad-
versely affect reliability. Although "unlimited" row accesses to the same row is allowed
within the refresh period; excessive row accesses to the same row over a long term can
result in degraded operation.
Table 75: Absolute Maximum Ratings
Symbol Parameter Min Max Unit Notes

VDD Voltage on VDD pin relative to VSS –0.4 1.5 V 1
VDDQ Voltage on VDDQ pin relative to VSS –0.4 1.5 V 1
VPP Voltage on VPP pin relative to VSS –0.4 3.0 V 3
VIN, VOUT Voltage on any pin relative to VSS –0.4 1.5 V
TSTG Storage temperature –55 150 °C 2
Notes: 1. VDD and VDDQ must be within 300mV of each other at all times, and VREF must not be
greater than 0.6 × VDDQ. When VDD and VDDQ are <500mV, VREF can be ื300mV.
2. Storage temperature is the case surface temperature on the center/top side of the
DRAM. For the measurement conditions, please refer to the JESD51-2 standard.
3. VPP must be equal to or greater than VDD/VDDQ at all times when powered.
DRAM Component Operating Temperature Range

Operating temperature, T OPER, is the case surface temperature on the center/top side of
the DRAM. For measurement conditions, refer to the JEDEC document JESD51-2.
Table 76: Temperature Range
Symbol Parameter Min Max Unit Notes

TOPER Normal operating temperature range –40 85 °C 1
Extended temperature range (optional) >85 125 °C 2
Notes: 1. The normal temperature range specifies the temperatures at which all DRAM specifica-
tions will be supported. During operation, the DRAM case temperature must be main-
tained between –40°C to 85°C under all operating conditions for the commercial offer-
ing.
2. Some applications require operation of the commercial and industrial temperature
DRAMs in the extended temperature range (between 85°C and 125°C case tempera-
ture). Full specifications are supported in this range, but the following additional condi-
tions apply:
• REFRESH commands must be doubled in frequency, reducing the refresh interval tREFI
to 3.9˩s. It is also possible to specify a component with 1X refresh (tREFI to 7.8˩s) in
the extended temperature range.
Electrical Characteristics – AC and DC Operating Conditions
• REFRESH command must be issued once every 0.975˩s if TC is greater than 105°C, once
every 1.95˩s if TC is greater than or equal to 95°C, once every 3.9˩s if TC is greater
than 85°C, and once every 7.8˩s if TC is less than 85°C.

Supply Operating Conditions
Table 77: Recommended Supply Operating Conditions
Rating
Symbol Parameter Min Typ Max Unit Notes
VDD Supply voltage 1.14 1.2 1.26 V 1, 2, 3, 4, 5
VDDQ Supply voltage for output 1.14 1.2 1.26 V 1, 2, 6
VPP Wordline supply voltage 2.375 2.5 2.750 V 7
Notes: 1. Under all conditions VDDQ must be less than or equal to VDD.
2. VDDQ tracks with VDD. AC parameters are measured with VDD and VDDQ tied together.
3. VDD slew rate between 300mV and 80% of VDD,min shall be between 0.004 V/ms and 600
V/ms, 20 MHz band-limited measurement.
4. VDD ramp time from 300mV to VDD,min shall be no longer than 200ms.
5. A stable valid VDD level is a set DC level (0 Hz to 250 KHz) and must be no less than
VDD,min and no greater than VDD,max. If the set DC level is altered anytime after initializa-
is final. AC noise of ±60mV (greater than 250 KHz) is allowed on VDD provided the noise
doesn't alter VDD to less than VDD,min or greater than VDD,max.
6. A stable valid VDDQ level is a set DC level (0 Hz to 250 KHz) and must be no less than
VDDQ,min and no greater than VDDQ,max. If the set DC level is altered anytime after initial-
ization, the DLL reset and calibrations must be performed again after the new set DC
level is final. AC noise of ±60mV (greater than 250 KHz) is allowed on VDDQ provided the
noise doesn't alter VDDQ to less than VDDQ,min or greater than VDDQ,max.
7. A stable valid VPP level is a set DC level (0 Hz to 250 KHz) and must be no less than
VPP,min and no greater than VPP,max. If the set DC level is altered anytime after initializa-
is final. AC noise of ±120mV (greater than 250 KHz) is allowed on VPP provided the noise
doesn't alter VPP to less than VPP,min or greater than VPP,max.
Table 78: VDD Slew Rate
Symbol Min Max Unit Notes

VDD_sl 0.004 600 V/ms 1, 2
VDD_on – 200 ms 3
Notes: 1. Measurement made between 300mV and 80% VDD (minimum level).
2. The DC bandwidth is limited to 20 MHz.
3. Maximum time to ramp VDD from 300 mV to VDD minimum.
Leakages
Table 79: Leakages
Condition Symbol Min Max Unit Notes

Input leakage (excluding ZQ and TEN) IIN –2 2 μA 1
ZQ leakage IZQ –50 10 μA 1
TEN leakage ITEN –6 10 μA 1, 2
VREFCA leakage IVREFCA –2 2 μA 3
Output leakage: VOUT = VDDQ IOZpd – 10 μA 4
Output leakage: VOUT = VSSQ IOZpu –50 – μA 4, 5
Notes: 1. Input under test 0V < VIN < 1.1V.

2. Additional leakage due to weak pull-down.
3. VREFCA = VDD/2, VDD at valid level after initialization.
4. DQs are disabled.
5. ODT is disabled with the ODT input HIGH.
VREFCA Supply
VREFCA is to be supplied to the DRAM and equal to V DD/2. The V REFCA is a reference sup-
ply input and therefore does not draw biasing current.
The DC-tolerance limits and AC-noise limits for the reference voltages V REFCA are illus-
trated in the figure below. The figure shows a valid reference voltage V REF(t) as a function
of time (VREF stands for V REFCA). V REF(DC) is the linear average of V REF(t) over a very long
period of time (1 second). This average has to meet the MIN/MAX requirements. Fur-
thermore, V REF(t) may temporarily deviate from V REF(DC) by no more than ±1% V DD for
the AC-noise limit.
Figure 205: VREFDQ Voltage Range
Voltage
VDD
VREF(t)
VREF AC-noise
VREF(DC) MAX
VREF(DC)
VDD/2
VREF(DC) MIN
VSS
Time
The voltage levels for setup and hold time measurements are dependent on V REF. V REF is
understood as V REF(DC), as defined in the above figure. This clarifies that DC-variations
of V REF affect the absolute voltage a signal has to reach to achieve a valid HIGH or LOW
level, and therefore, the time to which setup and hold is measured. System timing and
voltage budgets need to account for V REF(DC) deviations from the optimum position
within the data-eye of the input signals. This also clarifies that the DRAM setup/hold
specification and derating values need to include time and voltage associated with V REF
AC-noise. Timing and voltage effects due to AC-noise on V REF up to the specified limit
(±1% of V DD) are included in DRAM timings and their associated deratings.
VREFDQ Supply and Calibration Ranges

The device internally generates its own V REFDQ. DRAM internal V REFDQ specification pa-
rameters: voltage range, step size, V REF step time, V REF full step time, and V REF valid level
are used to help provide estimated values for the internal V REFDQ and are not pass/fail
limits. The voltage operating range specifies the minimum required range for DDR4
SDRAM devices. The minimum range is defined by V REFDQ,min and V REFDQ,max. A cali-
bration sequence should be performed by the DRAM controller to adjust V REFDQ and
optimize the timing and voltage margin of the DRAM data input receivers.
Table 80: VREFDQ Specification
Parameter Symbol Min Typ Max Unit Notes

Range 1 VREFDQ operating points VREFDQ R1 60% – 92% VDDQ 1, 2
Range 2 VREFDQ operating points VREFDQ R2 45% – 77% VDDQ 1, 2
VREF step size VREF,step 0.5% 0.65% 0.8% VDDQ 3
VREF set tolerance VREF,set_tol –1.625% 0% 1.625% VDDQ 4, 5, 6
–0.15% 0% 0.15% VDDQ 4, 7, 8
VREF step time VREF,time – – 150 ns 9, 10, 11
VREF valid tolerance VREF_val_tol –0.15% 0% 0.15% VDDQ 12
Notes: 1. VREF(DC) voltage is referenced to VDDQ(DC). VDDQ(DC) is 1.2V.

2. DRAM range 1 or range 2 is set by the MRS6[6]6.
3. VREF step size increment/decrement range. VREF at DC level.
4. VREF,new = VREF,old ±n × VREF,step; n = number of steps. If increment, use “+,” if decrement,
use “-.”
5. For n >4, the minimum value of VREF setting tolerance = VREF,new - 1.625% × VDDQ. The
maximum value of VREF setting tolerance = VREF,new + 1.625% × VDDQ.
6. Measured by recording the MIN and MAX values of the VREF output over the range,
drawing a straight line between those points, and comparing all other VREF output set-
tings to that line.
7. For n ื4, the minimum value of VREF setting tolerance = VREF,new - 0.15% × VDDQ. The
maximum value of VREF setting tolerance = VREF,new + 0.15% × VDDQ.
8. Measured by recording the MIN and MAX values of the VREF output across four consecu-
tive steps (n = 4), drawing a straight line between those points, and comparing all VREF
output settings to that line.
9. Time from MRS command to increment or decrement one step size for VREF.
10. Time from MRS command to increment or decrement more than one step size up to the
full range of VREF.
11. If the VREF monitor is enabled, VREF must be derated by +10ns if DQ bus load is 0pF and
an additional +15 ns/pF of DQ bus loading.
12. Only applicable for DRAM component-level test/characterization purposes. Not applica-
ble for normal mode of operation. VREF valid qualifies the step times, which will be char-
acterized at the component level.
VREFDQ Ranges
MR6[6] selects range 1 (60% to 92.5% of V DDQ) or range 2 (45% to 77.5% of V DDQ), and
MR6[5:0] sets the V REFDQ level, as listed in the following table. The values in MR6[6:0]
will update the V DDQ range and level independent of MR6[7] setting. It is recommended
MR6[7] be enabled when changing the settings in MR6[6:0], and it is highly recommen-
ded MR6[7] be enabled when changing the settings in MR6[6:0] multiple times during a
calibration routine.
Table 81: VREFDQ Range and Levels
MR6[6] 0 = MR6[6] 1 = MR6[6] 0 = MR6[6] 1 =

MR6[5:0] Range 1 Range 2 MR6[5:0] Range 1 Range 2
00 0000 60.00% 45.00% 01 1010 76.90% 61.90%
00 0001 60.65% 45.65% 01 1011 77.55% 62.55%
00 0010 61.30% 46.30% 01 1100 78.20% 63.20%
00 0011 61.95% 46.95% 01 1101 78.85% 63.85%
00 0100 62.60% 47.60% 01 1110 79.50% 64.50%
00 0101 63.25% 48.25% 01 1111 80.15% 65.15%
00 0110 63.90% 48.90% 10 0000 80.80% 65.80%
00 0111 64.55% 49.55% 10 0001 81.45% 66.45%
00 1000 65.20% 50.20% 10 0010 82.10% 67.10%
00 1001 65.85% 50.85% 10 0011 82.75% 67.75%
00 1010 66.50% 51.50% 10 0100 83.40% 68.40%
00 1011 67.15% 52.15% 10 0101 84.05% 69.05%
00 1100 67.80% 52.80% 10 0110 84.70% 69.70%
00 1101 68.45% 53.45% 10 0111 85.35% 70.35%
00 1110 69.10% 54.10% 10 1000 86.00% 71.00%
00 1111 69.75% 54.75% 10 1001 86.65% 71.65%
01 0000 70.40% 55.40% 10 1010 87.30% 72.30%
01 0001 71.05% 56.05% 10 1011 87.95% 72.95%
01 0010 71.70% 56.70% 10 1100 88.60% 73.60%
01 0011 72.35% 57.35% 10 1101 89.25% 74.25%
01 0100 73.00% 58.00% 10 1110 89.90% 74.90%
01 0101 73.65% 58.65% 10 1111 90.55% 75.55%
01 0110 74.30% 59.30% 11 0000 91.20% 76.20%
01 0111 74.95% 59.95% 11 0001 91.85% 76.85%
01 1000 75.60% 60.60% 11 0010 92.50% 77.50%
01 1001 76.25% 61.25% 11 0011 to 11 1111 are reserved
Electrical Characteristics – AC and DC Single-Ended Input
Measurement Levels
Electrical Characteristics – AC and DC Single-Ended Input Measurement

Levels
RESET_n Input Levels
Table 82: RESET_n Input Levels (CMOS)
Parameter Symbol Min Max Unit Note

AC input high voltage VIH(AC)_RESET 0.8 × VDD VDD V 1
DC input high voltage VIH(DC)_RESET 0.7 × VDD VDD V 2
DC input low voltage VIL(DC)_RESET VSS 0.3 × VDD V 3
AC input low voltage VIL(AC)_RESET VSS 0.2 × VDD V 4
Rising time tR_RESET – 1 μs 5
RESET pulse width after power-up tPW_RESET_S 1 – μs 6, 7
RESET pulse width during power-up tPW_RESET_L 200 – μs 6
Notes: 1. Overshoot should not exceed the VIN shown in the Absolute Maximum Ratings table.
2. After RESET_n is registered HIGH, the RESET_n level must be maintained above
VIH(DC)_RESET, otherwise operation will be uncertain until it is reset by asserting RESET_n
signal LOW.
3. After RESET_n is registered LOW, the RESET_n level must be maintained below VIL(DC)_RE-
t
SET during PW_RESET, otherwise the DRAM may not be reset.
4. Undershoot should not exceed the VIN shown in the Absolute Maximum Ratings table.
5. Slope reversal (ring-back) during this level transition from LOW to HIGH should be miti-
gated as much as possible.
6. RESET is destructive to data contents.
7. See RESET Procedure at Power Stable Condition figure.
Figure 206: RESET_n Input Slew Rate Definition

tPW_RESET
VIH(AC)_RESET,min
VIH(DC)_RESET,min
VIL(DC)_RESET,max
VIL(AC)_RESET,max
tR_RESET
Command/Address Input Levels
Table 83: Command and Address Input Levels: DDR4-1600 Through DDR4-2400

AC input high voltage VIH(AC) VREF + 100 VDD5 mV 1, 2, 3
DC input high voltage VIH(DC) VREF + 75 VDD mV 1, 2
Measurement Levels
Table 83: Command and Address Input Levels: DDR4-1600 Through DDR4-2400 (Continued)

DC input low voltage VIL(DC) VSS VREF - 75 mV 1, 2
AC input low voltage VIL(AC) VSS5 VREF - 100 mV 1, 2, 3
Reference voltage for CMD/ADDR inputs VREFFCA(DC) 0.49 × VDD 0.51 × VDD V 4
Notes: 1. For input except RESET_n. VREF = VREFCA(DC).

2. VREF = VREFCA(DC).
3. Input signal must meet VIL/VIH(AC) to meet tIS timings and VIL/VIH(DC) to meet tIH timings.
4. The AC peak noise on VREF may not allow VREF to deviate from VREFCA(DC) by more than
±1% VDD (for reference: approximately ±12mV).
5. Refer to “Overshoot and Undershoot Specifications.”
Table 84: Command and Address Input Levels: DDR4-2666


Table 85: Command and Address Input Levels: DDR4-2933 and DDR4-3200


Measurement Levels
Table 86: Single-Ended Input Slew Rates

Single-ended input slew rate – CA SRCA 1.0 7.0 V/ns 1, 2, 3, 4
Notes: 1. For input except RESET_n.

3. tIS/tIH timings assume SR
CA = 1V/ns.
4. Measured between VIH(AC) and VIL(AC) for falling edges and between VIL(AC) and VIH(AC)
for rising edges
Figure 207: Single-Ended Input Slew Rate Definition

TFse TRse
VIH(AC)
VIH(DC)
VREFCA
VIL(DC)
VIL(AC)
Command, Control, and Address Setup, Hold, and Derating

The total tIS (setup time) and tIH (hold time) required is calculated to account for slew
rate variation by adding the data sheet tIS (base) values, the V IL(AC)/VIH(AC) points, and
IL(DC)/VIH(DC) points; to the ˂ IS and ˂ IH derating values, re-
tIH (base) values, the V t t
spectively. The base values are derived with single-end signals at 1V/ns and differential
clock at 2 V/ns. Example: tIS (total setup time) = tIS (base) + ˂tIS. For a valid transition,
the input signal has to remain above/below V IH(AC)/VIL(AC) for the time defined by tVAC.
Although the total setup time for slow slew rates might be negative (for example, a valid
input signal will not have reached V IH(AC)/VIL(AC) at the time of the rising clock transi-
tion), a valid input signal is still required to complete the transition and to reach
VIH(AC)/VIL(AC). For slew rates that fall between the values listed in derating tables, the
derating values may be obtained by linear interpolation.
Setup (tIS) nominal slew rate for a rising signal is defined as the slew rate between the
last crossing of V IL(DC)max and the first crossing of V IH(AC)min that does not ring back be-
low V IH(DC)min . Setup (tIS) nominal slew rate for a falling signal is defined as the slew
Measurement Levels
rate between the last crossing of V IH(DC)min and the first crossing of V IL(AC)max that does
not ring back above V IL(DC)max.
Hold (tIH) nominal slew rate for a rising signal is defined as the slew rate between the
last crossing of V IL(DC)max and the first crossing of V IH(AC)min that does not ring back be-
low V IH(DC)min. Hold (tIH) nominal slew rate for a falling signal is defined as the slew rate
between the last crossing of V IH(DC)min and the first crossing of V IL(AC)minthat does not
ring back above V IL(DC)max.
Table 87: Command and Address Setup and Hold Values Referenced – AC/DC-Based
Symbol 1600 1866 2133 2400 2666 2933 3200 Unit Reference
tIS(base, AC100) 115 100 80 62 – – – ps VIH(AC)/VIL(AC)
tIH(base, DC75) 140 125 105 87 – – – ps VIH(DC)/VIL(DC)
tIS(base, AC90) – – – – 55 48 40 ps VIH(AC)/VIL(AC)
tIH(base, DC65) – – – – 80 73 65 ps VIH(DC)/VIL(DC)
tIS/tIH(Vref) 215 200 180 162 145 138 130 ps VIH(DC)/VIL(DC)
Table 88: Derating Values for tIS/tIH – AC100DC75-Based
˂tIS with AC100 Threshold, ˂tIH with DC75 Threshold Derating (ps) – AC/DC-Based
CMD/ CK, CK# Differential Slew Rate
ADDR 10.0 V/ns 8.0 V/ns 6.0 V/ns 4.0 V/ns 3.0 V/ns 2.0 V/ns 1.5 V/ns 1.0 V/ns
Slew Rate
V/ns ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIH ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH
7.0 76 54 76 55 77 56 79 58 82 60 86 64 94 73 111 89
6.0 73 53 74 53 75 54 77 56 79 58 83 63 92 71 108 88
5.0 70 50 71 51 72 52 74 54 76 56 80 60 88 68 105 85
4.0 65 46 66 47 67 48 69 50 71 52 75 56 83 65 100 81
3.0 57 40 57 41 58 42 60 44 63 46 67 50 75 58 92 75
2.0 40 28 41 28 42 29 44 31 46 33 50 38 58 46 75 63
1.5 23 15 24 16 25 17 27 19 29 21 33 25 42 33 58 50
1.0 –10 –10 –9 –9 –8 –8 –6 –6 –4 –4 0 0 8 8 25 25
0.9 –17 –14 –16 –14 –15 –13 –13 –10 –11 –8 –7 –4 1 4 18 21
0.8 –26 –19 –25 –19 –24 –18 –22 –16 –20 –14 –16 –9 –7 –1 9 16
0.7 –37 –26 –36 –25 –35 –24 –33 –22 –31 –20 –27 –16 –18 –8 –2 9
0.6 –52 –35 –51 –34 –50 –33 –48 –31 –46 –29 –42 –25 –33 –17 –17 0
0.5 –73 –48 –72 –47 –71 –46 –69 –44 –67 –42 –63 –38 –54 –29 –38 –13
0.4 –104 –66 –103 –66 –102 –65 –100 –63 –98 –60 –94 –56 –85 –48 –69 –31
Measurement Levels
Table 89: Derating Values for tIS/tIH – AC90/DC65-Based
˂tIS with AC90 Threshold, ˂tIH with DC65 Threshold Derating (ps) – AC/DC-Based
CMD/ CK, CK# Differential Slew Rate
ADDR 10.0 V/ns 8.0 V/ns 6.0 V/ns 4.0 V/ns 3.0 V/ns 2.0 V/ns 1.5 V/ns 1.0 V/ns
Slew Rate
V/ns ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH ˂tIH ˂tIH ˂tIS ˂tIH ˂tIS ˂tIH
7.0 68 47 69 47 70 48 72 50 73 52 77 56 85 63 100 78
6.0 66 45 67 46 68 47 69 49 71 50 75 54 83 62 98 77
5.0 63 43 64 44 65 45 66 46 68 48 72 52 80 60 95 75
4.0 59 40 59 40 60 41 62 43 64 45 68 49 75 56 90 71
3.0 51 34 52 35 53 36 54 38 56 40 60 43 68 51 83 66
2.0 36 24 37 24 38 25 39 27 41 29 45 33 53 40 68 55
1.5 21 13 22 13 23 14 24 16 26 18 30 22 38 29 53 44
1.0 –9 –9 –8 –8 –8 –8 –6 –6 –4 –4 0 0 8 8 23 23
0.9 –15 –13 –15 –12 –14 –11 –12 –9 –10 –7 –6 –4 1 4 16 19
0.8 –23 –17 –23 –17 –22 –16 –20 –14 –18 –12 –14 –8 –7 –1 8 14
0.7 –34 –23 –33 –22 –32 –21 –30 –20 –28 –18 –25 –14 –17 –6 –2 9
0.6 –47 –31 –47 –30 –46 –29 –44 –27 –42 –25 –38 –22 –31 –14 –16 1
0.5 –67 –42 –66 –41 –65 –40 –63 –38 –61 –36 –58 –33 –50 –25 –35 –10
0.4 –95 –58 –95 –57 –94 –56 –92 –54 –90 –53 –86 –49 –79 –41 –64 –26
Data Receiver Input Requirements

The following parameters apply to the data receiver Rx MASK operation detailed in the
Write Timing section, Data Strobe-to-Data Relationship.
The rising edge slew rates are defined by srr1 and srr2. The slew rate measurement
points for a rising edge are shown in the figure below. A LOW-to-HIGH transition time,
tr1, is measured from 0.5 × V diVW,max below V CENTDQ,midpoint to the last transition
through 0.5 × V diVW,max above V CENTDQ,midpoint; tr2 is measured from the last transition
through 0.5 × V diVW,max above V CENTDQ,midpoint to the first transition through the 0.5 ×
VIHL(AC)min above V CENTDQ,midpoint.
The falling edge slew rates are defined by srf1 and srf2. The slew rate measurement
points for a falling edge are shown in the figure below. A HIGH-to-LOW transition time,
tf1, is measured from 0.5 × V diVW,max above V CENTDQ,midpoint to the last transition
through 0.5 × V diVW,max below V CENTDQ,midpoint; tf2 is measured from the last transition
through 0.5 × V diVW,max below V CENTDQ,midpoint to the first transition through the 0.5 ×
VIHL(AC)min below V CENTDQ,midpoint.
Measurement Levels
Figure 208: DQ Slew Rate Definitions
tr2
VIHL(AC)min
0.5 ×
VIHL(AC)min
0.5 × VdiVW,max
VdiVW,max
VIHL(AC)min
0.5 × VdiVW,max
0.5 ×
tr1
tf1
VIHL(AC)min
0.5 ×
VIHL(AC)min
0.5 × VdiVW,max
VdiVW,max
VIHL(AC)min
0.5 × VdiVW,max
0.5 ×
tf2
Notes: 1. Rising edge slew rate equation srr1 = VdiVW,max/(tr1).

2. Rising edge slew rate equation srr2 = (VIHL(AC)min - VdiVW,max )/(2 × tr2).
3. Falling edge slew rate equation srf1 = VdiVW,max/(tf1).
4. Falling edge slew rate equation srf2 = (VIHL(AC)min - VdiVW,max )/(2 × tf2).
Table 90: DQ Input Receiver Specifications

DDR4-1600,
1866, 2133 DDR4-2400 DDR4-2666 DDR4-2933 DDR4-3200 Not
Parameter Symbol Min Max Min Max Min Max Min Max Min Max Unit es
VIN Rx mask input VdiVW – 136 – 130 – 120 – 115 – 110 mV 2, 3
peak-to-peak
DQ Rx input tim- TdiVW – 0.2 – 0.2 – 0.22 – 0.23 – 0.23 UI 2, 3
ing window
DQ AC input VIHL(AC) 186 – 160 – 150 – 145 – 140 – mV 4, 5
swing peak-to-
peak
Measurement Levels
Table 90: DQ Input Receiver Specifications (Continued)

DDR4-1600,
1866, 2133 DDR4-2400 DDR4-2666 DDR4-2933 DDR4-3200 Not
Parameter Symbol Min Max Min Max Min Max Min Max Min Max Unit es
DQ input pulse TdiPW 0.58 – 0.58 – 0.58 – 0.58 – 0.58 – UI 6
width
DQS-to-DQ Rx tDQS2D –0.17 0.17 –0.17 0.17 –0.19 0.19 –0.22 0.22 –0.22 0.22 UI 7
mask offset Q
DQ-to-DQ Rx mask tDQ2DQ – 0.1 – 0.1 – 0.105 – 0.115 – 0.125 UI 8
offset
Input slew rate srr1, srf1 1 9 1 9 1 9 1 9 1 9 V/ns 9
over VdiVW if tCK ุ
0.937ns
Input slew rate srr1, srf1 – – 1.25 9 1.25 9 1.25 9 1.25 9 V/ns 9
over VdiVW if
0.937ns > tCK ุ
0.625ns
Rising input slew srr2 0.2 × 9 0.2 × 9 0.2 × 9 0.2 × 9 0.2 × 9 V/ns 10
rate over 1/2 srr1 srr1 srr1 srr1 srr1
VIHL(AC)
Falling input slew srf2 0.2 × 9 0.2 × 9 0.2 × 9 0.2 × 9 0.2 × 9 V/ns 10
rate over 1/2 srf1 srf1 srf1 srf1 srf1
VIHL(AC)
Notes: 1. All Rx mask specifications must be satisfied for each UI. For example, if the minimum in-
put pulse width is violated when satisfying TdiVW (MIN), VdiVW,max, and minimum slew
rate limits, then either TdiVW (MIN) or minimum slew rates would have to be increased
to the point where the minimum input pulse width would no longer be violated.
2. Data Rx mask voltage and timing total input valid window where VdiVW is centered
around VCENTDQ,midpoint after VREFDQ training is completed. The data Rx mask is applied
per bit and should include voltage and temperature drift terms. The input buffer design
specification is to achieve at least a BER =1e- 16 when the Rx mask is not violated.
3. Defined over the DQ internal VREF range 1.
4. Overshoot and undershoot specifications apply.
5. DQ input pulse signal swing into the receiver must meet or exceed VIHL(AC)min. VIHL(AC)min
is to be achieved on an UI basis when a rising and falling edge occur in the same UI (a
valid TdiPW).
6. DQ minimum input pulse width defined at the VCENTDQ,midpoint.
7. DQS-to-DQ Rx mask offset is skew between DQS and DQ within a nibble (x4) or word
(x8, x16 [for x16, the upper and lower bytes are treated as separate x8s]) at the SDRAM
balls over process, voltage, and temperature.
8. DQ-to-DQ Rx mask offset is skew between DQs within a nibble (x4) or word (x8, x16) at
the SDRAM balls for a given component over process, voltage, and temperature.
9. Input slew rate over VdiVW mask centered at VCENTDQ,midpoint. Slowest DQ slew rate to
fastest DQ slew rate per transition edge must be within 1.7V/ns of each other.
10. Input slew rate between VdiVW mask edge and VIHL(AC)min points.
The following figure shows the Rx mask relationship to the input timing specifications
relative to system tDS and tDH. The classical definition for tDS/tDH required a DQ rising
Measurement Levels
and falling edges to not violate tDS and tDH relative to the DQS strobe at any time; how-
ever, with the Rx mask tDS and tDH can shift relative to the DQS strobe provided the
input pulse width specification is satisfied and the Rx mask is not violated.
Figure 209: Rx Mask Relative to tDS/tDH
TdiPW
VIH(DC)
0.5 × VdiVW
VdiVW
Rx VCENTDQ,pin mean
Mask
0.5 × VdiVW
VIL(DC)
tf1 TdiVW tr1
tDS= Greater of 0.5 × TdiVW tDH= Greater of 0.5 × TdiVW

or or
0.5 × (TdiPW + VdiVW/tf1) 0.5 × (TdiPW + VdiVW/tr1)
DQS_c
DQS_t
The following figure and table show an example of the worst case Rx mask required if
the DQS and DQ pins do not have DRAM controller to DRAM write DQ training. The
figure and table show that without DRAM write DQ training, the Rx mask would in-
crease from 0.2UI to essentially 0.54UI. This would also be the minimum tDS and tDH
required as well.
Measurement Levels
Figure 210: Rx Mask Without Write Training
TdiVW + 2 × tDQS2DQ
VIH(DC)
0.5 × VdiVW
VdiVW
0.5 × VdiVW
VIL(DC)
tDS tDH
0.5 × TdiVW + tDQS2DQ 0.5 × TdiVW + tDQS2DQ
DQS_c
DQS_t
Table 91: Rx Mask and tDS/tDH without Write Training
Rx Mask
VIHL(AC) TdiPW VdiVW TdiVW tDQS2DQ tDQ2DQ with Write tDS + tDH
DDR4
(mV) (UI) (mV) (UI) (UI) (UI) Train (ps)
(ps)
1600 186 0.58 136 0.2 ±0.17 0.1 125 338
1866 186 0.58 136 0.2 ±0.17 0.1 107.1 289
2133 186 0.58 136 0.2 ±0.17 0.1 94 253
2400 160 0.58 130 0.2 ±0.17 0.1 83.3 225
2666 150 0.58 120 0.22 ±0.19 0.105 82.5 225
2933 145 0.58 115 0.23 ±0.22 0.115 78.4 228
3200 140 0.58 110 0.23 ±0.22 0.125 71.8 209
Note: 1. VIHL(AC), VdiVW, and VILH(DC) referenced to VCENTDQ,midpoint.
Connectivity Test (CT) Mode Input Levels
Table 92: TEN Input Levels (CMOS)

TEN AC input high voltage VIH(AC)_TEN 0.8 × VDD VDD V 1
TEN DC input high voltage VIH(DC)_TEN 0.7 × VDD VDD V
TEN DC input low voltage VIL(DC)_TEN VSS 0.3 × VDD V
TEN AC input low voltage VIL(AC)_TEN VSS 0.2 × VDD V 2
TEN falling time tF_TEN – 10 ns
Measurement Levels
Table 92: TEN Input Levels (CMOS) (Continued)

TEN rising time tR_TEN – 10 ns
Notes: 1. Overshoot should not exceed the VIN values in the Absolute Maximum Ratings table.
2. Undershoot should not exceed the VIN values in the Absolute Maximum Ratings table.
Figure 211: TEN Input Slew Rate Definition
VIH(AC)_TENmin
VIH(DC)_TENmin
VIL(DC)_TENmin
VIL(AC)_TENmin
tF_TEN tR_TEN
Table 93: CT Type-A Input Levels

CTipA AC input high voltage VIH(AC) VREF + 200 VDD11 V 2, 3
CTipA DC input high voltage VIH(DC) VREF + 150 VDD V 2, 3
CTipA DC input low voltage VIL(DC) VSS VREF - 150 V 2, 3
CTipA AC input low voltage VIL(AC) VSS1 1 VREF - 200 V 2, 3
CTipA falling time tF_CTipA – 5 ns 2
CTipA rising time tR_CTipA – 5 ns 2
Notes: 1. Refer to Overshoot and Undershoot Specifications.

2. CT Type-A inputs: CS_n, BG[1:0], BA[1:0], A[9:0], A10/AP, A11, A12/BC_n, A13, WE_n/A14,
CAS_n/A15, RAS_n/A16, A17, CKE, ACT_n, ODT, CLK_t, CLK_C, PAR.
3. VREFCA = 0.5 × VDD.
Figure 212: CT Type-A Input Slew Rate Definition
VIH(AC)_CTipAmin
VIH(DC)_CTipAmin
VREFCA
VIL(DC)_CTipAmax
VIL(AC)_CTipAmax
tF_CTipA tR_CTipA
Measurement Levels
Table 94: CT Type-B Input Levels

CTipB AC input high voltage VIH(AC) VREF + 300 VDD11 V 2, 3
CTipB DC input high voltage VIH(DC) VREF + 200 VDD V 2, 3
CTipB DC input low voltage VIL(DC) VSS VREF - 200 V 2, 3
CTipB AC input low voltage VIL(AC) VSS1 1 VREF - 300 V 2, 3
CTipB falling time tF_CTipB – 5 ns 2
CTipB rising time tR_CTipB – 5 ns 2

2. CT Type-B inputs: DML_n/DBIL_n, DMU_n/DBIU_n and DM_n/DBI_n.
3. VREFDQ should be 0.5 × VDD
Figure 213: CT Type-B Input Slew Rate Definition
VIH(AC)_CTipBmin
VIH(DC)_CTipBmin
VREFDQ
VIL(DC)_CTipBmax
VIL(AC)_CTipBmax
tF_CTipB tR_CTipB
Table 95: CT Type-C Input Levels (CMOS)

CTipC AC input high voltage VIH(AC)_CTipC 0.8 × VDD VDD 1 V 2
CTipC DC input high voltage VIH(DC)_CTipC 0.7 × VDD VDD V 2
CTipC DC input low voltage VIL(DC)_CTipC VSS 0.3 × VDD V 2
CTipC AC input low voltage VIL(AC)_CTipC VSS1 0.2 × VDD V 2
CTipC falling time tF_CTipC – 10 ns 2
CTipC rising time tR_CTipC – 10 ns 2

2. CT Type-C inputs: Alert_n.
Measurement Levels
Figure 214: CT Type-C Input Slew Rate Definition
VIH(AC)_TENmin
VIH(DC)_TENmin
VIL(DC)_TENmin
VIL(AC)_TENmin
tF_TEN tR_TEN
Table 96: CT Type-D Input Levels

CTipD AC input high voltage VIH(AC)_CTipD 0.8 × VDD VDD V 4
CTipD DC input high voltage VIH(DC)_CTipD 0.7 × VDD VDD V 2
CTipD DC input low voltage VIL(DC)_CTipD VSS 0.3 × VDD V 1
CTipD AC input low voltage VIL(AC)_CTipD VSS 0.2 × VDD V 5
Rising time tR_RESET – 1 μs 3
RESET pulse width - after power-up tPW_RESET_S 1 – μs
RESET pulse width - during power-up tPW_RESET_L 200 – μs
Notes: 1. After RESET_n is registered LOW, the RESET_n level must be maintained below VIL(DC)_RE-
t
SET during PW_RESET, otherwise, the DRAM may not be reset.
2. After RESET_n is registered HIGH, the RESET_n level must be maintained above
VIH(DC)_RESET, otherwise, operation will be uncertain until it is reset by asserting RESET_n
signal LOW.
3. Slope reversal (ring-back) during this level transition from LOW to HIGH should be miti-
gated as much as possible.
4. Overshoot should not exceed the VIN values in the Absolute Maximum Ratings table.
5. Undershoot should not exceed the VIN values in the Absolute Maximum Ratings table.
6. CT Type-D inputs: RESET_n; same requirements as in normal mode.
Figure 215: CT Type-D Input Slew Rate Definition

tPW_RESET
VIH(AC)_RESETmin
VIH(DC)_RESETmin
VIL(DC)_RESETmax
VIL(AC)_RESETmax
tR_RESET
Electrical Characteristics – AC and DC Differential Input Meas-
urement Levels
Electrical Characteristics – AC and DC Differential Input Measurement

Levels
Differential Inputs
Figure 216: Differential AC Swing and “Time Exceeding AC-Level” tDVAC
tDVAC
VIH,diff(AC)min
VIH,diff,min
CK_t, CK_c
0.0
VIL,diff,max
VIL,diff(AC)max
Half cycle tDVAC
Notes: 1. Differential signal rising edge from VIL,diff,max to VIH,diff(AC)min must be monotonic slope.
2. Differential signal falling edge from IH,diff,min to VIL,diff(AC)max must be monotonic slope.
Table 97: Differential Input Swing Requirements for CK_t, CK_c
DDR4-1600 / DDR4-2400 /
Sym- 1866 / 2133 2666 DDR4-2933 DDR4-3200 Note
Parameter bol Min Max Min Max Min Max Min Max Unit s
Differential input high VIHdiff 150 Note 3 135 Note 3 125 Note 3 110 Note 3 mV 1
Differential input low VILdiff Note 3 –150 Note 3 -135 Note 3 -125 Note 3 -110 mV 1
Differential input high VIH- 2× Note 3 2× Note 3 2× Note 3 2× Note 3 V 2
(AC) diff(AC) (VIH(AC) (VIH(AC) (VIH(AC) (VIH(AC)
- VREF) - VREF) - VREF) - VREF)
urement Levels
Table 97: Differential Input Swing Requirements for CK_t, CK_c (Continued)
DDR4-1600 / DDR4-2400 /
Sym- 1866 / 2133 2666 DDR4-2933 DDR4-3200 Note
Parameter bol Min Max Min Max Min Max Min Max Unit s
Differential input low VIL- Note 3 2× Note 3 2× Note 3 2× Note 3 2× V 2
(AC) diff(AC) (VIL(AC) - (VIL(AC) - (VIL(AC) - (VIL(AC) -
VREF) VREF) VREF) VREF)
Notes: 1. Used to define a differential signal slew-rate.

2. For CK_t, CK_c use VIH(AC) and VIL(AC) of ADD/CMD and VREFCA.
3. These values are not defined; however, the differential signals (CK_t, CK_c) need to be
within the respective limits, VIH(DC)max and VIL(DC)min for single-ended signals as well as
the limitations for overshoot and undershoot.
Table 98: Minimum Time AC Time tDVAC for CK

tDVAC (ps) at |VIH,diff(AC) to VIL,diff(AC)|
Slew Rate (V/ns) 200mV TBDmV
>4.0 120 TBD
4.0 115 TBD
3.0 110 TBD
2.0 105 TBD
1.9 100 TBD
1.6 95 TBD
1.4 90 TBD
1.2 85 TBD
1.0 80 TBD
<1.0 80 TBD
Note: 1. Below VIL(AC).
Single-Ended Requirements for CK Differential Signals

Each individual component of a differential signal (CK_t, CK_c) has to comply with cer-
tain requirements for single-ended signals. CK_t and CK_c have to reach approximately
VSEHmin/VSEL,max, which are approximately equal to the AC levels V IH(AC) and V IL(AC) for
ADD/CMD signals in every half-cycle. The applicable AC levels for ADD/CMD might
differ per speed-bin, and so on. For example, if a value other than 100mV is used for
ADD/CMD V IH(AC) and V IL(AC) signals, then these AC levels also apply for the single-
ended signals CK_t and CK_c.
While ADD/CMD signal requirements are with respect to V REFCA, the single-ended com-
ponents of differential signals have a requirement with respect to V DD/2; this is nomi-
nally the same. The transition of single-ended signals through the AC levels is used to
measure setup time. For single-ended components of differential signals the require-
ment to reach V SEL,max/VSEH,min has no bearing on timing, but adds a restriction on the
common mode characteristics of these signals.
urement Levels
Figure 217: Single-Ended Requirements for CK
VDD or VDDQ
VSEH,min
VDD/2 or VDDQ/2
VSEH
CK
VSEL,max
VSEL
VSS or VSSQ
Table 99: Single-Ended Requirements for CK
DDR4-1600 /
1866 / 2133 DDR4-2400 / 2666 DDR4-2933 / 3200
Parameter Symbol Min Max Min Max Unit Notes
Single-ended high level for VSEH VDD/2 + Note 3 VDD/2 + Note 3 VDD/2 + Note 3 V 1, 2
CK_t, CK_c 0.100 0.095 0.085
Single-ended low level for VSEL Note 3 VDD/2 - Note 3 VDD/2 - Note 3 VDD/2 - V 1, 2
CK_t, CK_c 0.100 0.095 0.085
Notes: 1. For CK_t, CK_c use VIH(AC) and VIL(AC) of ADD/CMD and VREFCA.
2. ADDR/CMD VIH(AC) and VIL(AC) based on VREFCA.
3. These values are not defined; however, the differential signal (CK_t, CK_c) need to be
within the respective limits, VIH(DC)max and VIL(DC)min for single-ended signals as well as
the limitations for overshoot and undershoot.
Slew Rate Definitions for CK Differential Input Signals
Table 100: CK Differential Input Slew Rate Definition
Measured
Description From To Defined by
Differential input slew rate for rising edge VIL,diff,max VIH,diff,min |VIH,diff,min - VIL,diff,max_˂TRdiff
Differential input slew rate for falling edge VIH,diff,min VIL,diff,max |VIH,diff,min - VIL,diff,max_˂TFdiff
Note: 1. The differential signal CK_t, CK_c must be monotonic between these thresholds.
urement Levels
Figure 218: Differential Input Slew Rate Definition for CK_t, CK_c
TRdiff
VIH,diff,min
CK Differential Input Voltage
VIL,diff,max
TFdiff
CK Differential Input Cross Point Voltage

To guarantee tight setup and hold times as well as output skew parameters with respect
to clock and strobe, each cross point voltage of differential input signal CK_t, CK_c must
meet the requirements shown below. The differential input cross point voltage V IX(CK) is
measured from the actual cross point of true and complement signals to the midlevel
between V DD and V SS.
Figure 219: VIX(CK) Definition

VDD
CK_c
VIX(CK)
VDD/2
VIX(CK) VIX(CK)
CK_t
VSEH VSEL
VSS
urement Levels
Table 101: Cross Point Voltage For CK Differential Input Signals at DDR4-1600 through DDR4-2400
DDR4-1600, 1866, 2133, 2400

Parameter Sym Input Level Min Max
Differential input VIX(CK) VSEH > VDD/2 + 145mV N/A 120mV
cross point volt- VDD/2 + 100mV ื VSEH ื VDD/2 + 145mV N/A (VSEH - VDD/2) - 25mV
age relative to
VDD/2 - 145mV ื VSEL ื VDD/2 - 100mV –(VDD/2 - VSEL) + 25mV N/A
VDD/2 for CK_t,
CK_c VSEL < VDD/2 - 145mV –120mV N/A
Table 102: Cross Point Voltage For CK Differential Input Signals at DDR4-2666 through DDR4-3200
DDR4-2666, 2933, 3200

Parameter Sym Input Level Min Max
Differential input VIX(CK) VSEH > VDD/2 + 145mV N/A 110mV
cross point volt- VDD/2 + 90mV ื VSEH ื VDD/2 + 145mV N/A (VSEH - VDD/2) - 30mV
age relative to
VDD/2 - 145mV ื VSEL ื VDD/2 - 90mV –(VDD/2 - VSEL) + 30mV N/A
VDD/2 for CK_t,
CK_c VSEL < VDD/2 - 145mV –110mV N/A
DQS Differential Input Signal Definition and Swing Requirements

Figure 220: Differential Input Signal Definition for DQS_t, DQS_c
DQS_t, DQS_c: Differential Input Voltage
VIH,diff,peak
Half cycle
0.0V
Half cycle
VIL,diff,peak
Table 103: DDR4-1600 through DDR4-2400 Differential Input Swing Requirements for DQS_t, DQS_c
DDR4-1600, 1866,
2133 DDR4-2400
Peak differential input high voltage VIH,diff,peak 186 VDDQ 160 VDDQ mV 1,2
urement Levels
(Continued)
DDR4-1600, 1866,
2133 DDR4-2400
Peak differential input low voltage VIL,diff,peak VSSQ –186 VSSQ –160 mV 1,2
Notes: 1. Minimum and maximum limits are relative to single-ended portion and can be exceeded
within allowed overshoot and undershoot limits.
2. Minimum value point is used to determine differential signal slew-rate.
DDR4-2666 DDR4-2933 DDR4-3200

Parameter Symbol Min Max Min Max Min Max Unit Notes
Peak differential input high volt- VIH,diff,peak 150 VDDQ 145 VDDQ 140 VDDQ mV 1,2
age
Peak differential input low volt- VIL,diff,peak VSSQ –150 VSSQ –145 VSSQ –140 mV 1,2
age
2. Minimum value point is used to determine differential signal slew-rate.
The peak voltage of the DQS signals are calculated using the following equations:
VIH,dif,Peak voltage = MAX(ft)
VIL,dif,Peak voltage = MIN(ft)
(ft) = DQS_t, DQS_c.
The MAX(f(t)) or MIN(f(t)) used to determine the midpoint from which to reference the
±35% window of the exempt non-monotonic signaling shall be the smallest peak volt-
age observed in all UIs.
Figure 221: DQS_t, DQS_c Input Peak Voltage Calculation and Range of Exempt non-Monotonic Sig-
naling
DQS_t, DQS_c: Single-Ended Input Voltages
DQS_t
+50%
+35%
MIN(ft) MAX(ft)
–35%
–50%
DQS_c
urement Levels
DQS Differential Input Cross Point Voltage

To achieve tight RxMask input requirements as well as output skew parameters with re-
spect to strobe, the cross point voltage of differential input signals (DQS_t, DQS_c) must
meet V IX_DQS,ratio in the table below. The differential input cross point voltage V IX_DQS
(VIX_DQS_FR and V IX_DQS_RF) is measured from the actual cross point of DQS_t, DQS_c
relative to the V DQS,mid of the DQS_t and DQS_c signals.
VDQS,mid is the midpoint of the minimum levels achieved by the transitioning DQS_t
and DQS_c signals, and noted by V DQS_trans. V DQS_trans is the difference between the low-
est horizontal tangent above V DQS,mid of the transitioning DQS signals and the highest
horizontal tangent below V DQS,mid of the transitioning DQS signals. A non-monotonic
transitioning signal’s ledge is exempt or not used in determination of a horizontal tan-
gent provided the said ledge occurs within ±35% of the midpoint of either V IH.DIFF.Peak
voltage (DQS_t rising) or V IL.DIFF.Peak voltage (DQS_c rising), as shown in the figure be-
low.
A secondary horizontal tangent resulting from a ring-back transition is also exempt in
determination of a horizontal tangent. That is, a falling transition’s horizontal tangent is
derived from its negative slope to zero slope transition (point A in the figure below), and
a ring-back’s horizontal tangent is derived from its positive slope to zero slope transi-
tion (point B in the figure below) and is not a valid horizontal tangent; a rising transi-
tion’s horizontal tangent is derived from its positive slope to zero slope transition (point
C in the figure below), and a ring-back’s horizontal tangent derived from its negative
slope to zero slope transition (point D in the figure below) and is not a valid horizontal
tangent.
Figure 222: VIXDQS Definition

Lowest horizontal tanget above VDQS,mid
of the transitioning signals
C
DQS_t
DQS_t, DQS_c: Single-Ended Input Voltages
VDQS_trans
VIX_DQS,FR VIX_DQS,RF
VDQS,mid
VDQS_trans/2
VIX_DQS,RF VIX_DQS,FR
DQS_c
Highest horizontal tanget below VDQS,mid

of the transitioning signals
VSSQ
Table 105: Cross Point Voltage For Differential Input Signals DQS
DDR4-1600, 1866, 2133, 2400,

2666, 2933, 3200
Parameter Symbol Min Max Unit Notes
DQS_t and DQS_c crossing relative to the VIX_DQS,ratio – 25 % 1, 2
midpoint of the DQS_t and DQS_c signal
swings
urement Levels
Table 105: Cross Point Voltage For Differential Input Signals DQS (Continued)
DDR4-1600, 1866, 2133, 2400,

2666, 2933, 3200
Parameter Symbol Min Max Unit Notes
VDQS,mid to Vcent(midpoint) offset VDQS,mid_to_Vcent – Note 3 mV 2
Notes: 1. VIX_DQS,ratio is DQS VIX crossing (VIX_DQS,FR or VIX_DQS,RF) divided by VDQS_trans. VDQS_trans is
the difference between the lowest horizontal tangent above VDQS,midd of the transition-
ing DQS signals and the highest horizontal tangent below VDQS,mid of the transitioning
DQS signals.
2. VDQS,mid will be similar to the VREFDQ internal setting value (Vcent(midpoint) offset) ob-
tained during VREF Training if the DQS and DQs drivers and paths are matched.
3. The maximum limit shall not exceed the smaller of VIH,diff,DQS minimum limit or 50mV.
Slew Rate Definitions for DQS Differential Input Signals
Table 106: DQS Differential Input Slew Rate Definition
Measured
Differential input slew rate for rising edge V IL,diff,DQS V IH,diff,DQS |VIH,diff,DQS - VIL,diff,DQS_˂TRdiff
Differential input slew rate for falling edge V IH,diff,DQS V IL,diff,DQS |VIHdiffDQS - VIL,diff,DQS_˂TFdiff
Note: 1. The differential signal DQS_t, DQS_c must be monotonic between these thresholds.
Figure 223: Differential Input Slew Rate and Input Level Definition for DQS_t, DQS_c
DQS_t, DQS_c: Differential Input Voltage
VIH,diff,peak
VIH,diff,DQS
0.0V
VIL,diff,DQS
TFdiff TRdiff
VIL,diff,peak
Table 107: DDR4-1600 through DDR4-2400 Differential Input Slew Rate and Input Levels for DQS_t,
DQS_c
DDR4-1600, 1866, 2133 DDR4-2400

Peak differential input high voltage VIH,diff,peak 186 VDDQ 160 VDDQ mV 1
Differential input high voltage VIH,diff,DQS 136 – 130 – mV 2, 3
Differential input low voltage VIL,diff,DQS – –136 – –130 mV 2, 3
urement Levels
DQS_c (Continued)
DDR4-1600, 1866, 2133 DDR4-2400

Peak differential input low voltage VIL,diff,peak VSSQ –186 VSSQ –160 mV 1
DQS differential input slew rate SRIdiff 3.0 18 3.0 18 V/ns 4, 5
2. Differential signal rising edge from VIL,diff,DQS to VIH,diff,DQS must be monotonic slope.
3. Differential signal falling edge from VIH,diff,DQS to VIL,diff,DQS must be monotonic slope.
4. Differential input slew rate for rising edge from VIL,diff,DQS to VIH,diff,DQS is defined by |
VIL,diff,min - VIH,diff,max_˂TRdiff.
5. Differential input slew rate for falling edge from VIH,diff,DQS to VIL,diff,DQS is defined by |
VIL,diff,min - VIH,diff,max_˂TFdiff.
DQS_c
DDR4-2666 DDR4-2933 DDR4-3200

Peak differential input VIH,diff,peak 150 VDDQ 145 VDDQ 140 VDDQ mV 1
high voltage
Differential input high VIH,diff,DQS 130 – 115 – 110 – mV 2, 3
voltage
Differential input low VIL,diff,DQS – –130 – –115 – –110 mV 2, 3
voltage
Peak differential input VIL,diff,peak VSSQ –150 VSSQ –145 VSSQ –140 mV 1
low voltage
DQS differential input SRIdiff 2.5 18 2.5 18 2.5 18 V/ns 4, 5
slew rate
2. Differential signal rising edge from VIL,diff,DQS to VIH,diff,DQS must be monotonic slope.
3. Differential signal falling edge from VIH,diff,DQS to VIL,diff,DQS must be monotonic slope.
4. Differential input slew rate for rising edge from VIL,diff,DQS to VIH,diff,DQS is defined by |
VIL,diff,min - VIH,diff,max_˂TRdiff.
5. Differential input slew rate for falling edge from VIH,diff,DQS to VIL,diff,DQS is defined by |
VIL,diff,min - VIH,diff,max_˂TFdiff.
Electrical Characteristics – Overshoot and Undershoot Specifi-
cations
Electrical Characteristics – Overshoot and Undershoot Specifications

Address, Command, and Control Overshoot and Undershoot Specifications
Table 109: ADDR, CMD, CNTL Overshoot and Undershoot/Specifications
DDR4- DDR4- DDR4- DDR4- DDR4- DDR4- DDR4-

Description 1600 1866 2133 2400 2666 2933 3200 Unit
Address and control pins (A[17:0], BG[1:0], BA[1:0], CS_n, RAS_n, CAS_n, WE_n, CKE, ODT, C2-0)
Area A: Maximum peak amplitude above VDD 0.06 0.06 0.06 0.06 0.06 0.06 0.06 V
absolute MAX
Area B: Amplitude allowed between VDD and 0.24 0.24 0.24 0.24 0.24 0.24 0.24 V
VDD absolute MAX
Area C: Maximum peak amplitude allowed for 0.30 0.30 0.30 0.30 0.30 0.30 0.30 V
undershoot below VSS
Area A maximum overshoot area per 1tCK 0.0083 0.0071 0.0062 0.0055 0.0055 0.0055 0.0055 V/ns
Area B maximum overshoot area per 1tCK 0.2550 0.2185 0.1914 0.1699 0.1699 0.1699 0.1699 V/ns
Area C maximum undershoot area per 1tCK 0.2644 0.2265 0.1984 0.1762 0.1762 0.1762 0.1762 V/ns
Figure 224: ADDR, CMD, CNTL Overshoot and Undershoot Definition

Absolute MAX overshoot
A Overshoot area above VDD absolute MAX
VDD absolute MAX
B Overshoot area below VDD absolute MAX

and above VDD MAX
Volts (V)
VDD
1tCK
VSS
C Undershoot area below VSS
Electrical Characteristics – Overshoot and Undershoot Specifi-
cations
Clock Overshoot and Undershoot Specifications
Table 110: CK Overshoot and Undershoot/ Specifications

CLK_t, CLK_n
Area A: Maximum peak amplitude above VDD 0.06 0.06 0.06 0.06 0.06 0.06 0.06 V
absolute MAX
Area B: Amplitude allowed between VDD and 0.24 0.24 0.24 0.24 0.24 0.24 0.24 V
VDD absolute MAX
undershoot below VSS
Area A maximum overshoot area per 1UI 0.0038 0.0032 0.0028 0.0025 0.0025 0.0025 0.0025 V/ns
Area B maximum overshoot area per 1UI 0.1125 0.0964 0.0844 0.0750 0.0750 0.0750 0.0750 V/ns
Area C maximum undershoot area per 1UI 0.1144 0.0980 0.0858 0.0762 0.0762 0.0762 0.0762 V/ns
Figure 225: CK Overshoot and Undershoot Definition

A Overshoot area above VDD absolute MAX
VDD absolute MAX
B Overshoot area below VDD absolute MAX

and above VDD MAX
Volts (V)
VDD
1UI
VSS
C Undershoot area below VSS
Electrical Characteristics – AC and DC Output Measurement
Levels
Data, Strobe, and Mask Overshoot and Undershoot Specifications
Table 111: Data, Strobe, and Mask Overshoot and Undershoot/ Specifications

DQS_t, DQS_n, LDQS_t, LDQS_n, UDQS_t, UDQS_n, DQ[0:15], DM/DBI, UDM/UDBI, LDM/LDBI,
Area A: Maximum peak amplitude above VDDQ 0.16 0.16 0.16 0.16 0.16 0.16 0.16 V
absolute MAX
Area B: Amplitude allowed between VDDQ and 0.24 0.24 0.24 0.24 0.24 0.24 0.24 V
VDDQ absolute MAX
undershoot below VSSQ
Area D: Maximum peak amplitude below VSSQ 0.10 0.10 0.10 0.10 0.10 0.10 0.10 V
absolute MIN
Area A maximum overshoot area per 1UI 0.0150 0.0129 0.0113 0.0100 0.0100 0.0100 0.0100 V/ns
Area B maximum overshoot area per 1UI 0.1050 0.0900 0.0788 0.0700 0.0700 0.0700 0.0700 V/ns
Area C maximum undershoot area per 1UI 0.1050 0.0900 0.0788 0.0700 0.0700 0.0700 0.0700 V/ns
Area D maximum undershoot area per 1UI 0.0150 0.0129 0.0113 0.0100 0.0100 0.0100 0.0100 V/ns
Figure 226: Data, Strobe, and Mask Overshoot and Undershoot Definition
A Overshoot area above VDDQ absolute MAX
VDDQ absolute MAX
B Overshoot area below VDDQ absolute MAX

and above VDDQ MAX
Volts (V)
VDDQ
1UI
VSSQ
C Undershoot area below VSSQ MIN and

above VSSQ absolute MIN
VSSQ absolute MIN
D Undershoot area below VSSQ absolute MIN
Absolute MAX undershoot
Electrical Characteristics – AC and DC Output Measurement Levels

Single-Ended Outputs
Table 112: Single-Ended Output Levels
Parameter Symbol DDR4-1600 to DDR4-3200 Unit

DC output high measurement level (for IV curve linearity) VOH(DC) 1.1 × VDDQ V
DC output mid measurement level (for IV curve linearity) VOM(DC) 0.8 × VDDQ V
DC output low measurement level (for IV curve linearity) VOL(DC) 0.5 × VDDQ V
AC output high measurement level (for output slew rate) VOH(AC) (0.7 + 0.15) × VDDQ V
Levels
Table 112: Single-Ended Output Levels (Continued)

AC output low measurement level (for output slew rate) VOL(AC) (0.7 - 0.15) × VDDQ V
Note: 1. The swing of ±0.15 × VDDQ is based on approximately 50% of the static single-ended
output peak-to-peak swing with a driver impedance of RZQ/7 and an effective test load
of 50˖ to VTT = VDDQ.
Using the same reference load used for timing measurements, output slew rate for fall-
ing and rising edges is defined and measured between V OL(AC) and V OH(AC) for single-
ended signals.
Table 113: Single-Ended Output Slew Rate Definition
Measured
Single-ended output slew rate for rising edge VOL(AC) VOH(AC) [VOH(AC) - VOL(AC)@˂TRse
Single-ended output slew rate for falling edge VOH(AC) VOL(AC) [VOH(AC) - VOL(AC)@˂TFse
Figure 227: Single-ended Output Slew Rate Definition

TRse
VOH(AC)
Single-Ended Output Voltage (DQ)
VOL(AC)
TFse
Levels
Table 114: Single-Ended Output Slew Rate

For RON = RZQ/7
DDR4-1600/ 1866 / 2133 /
2400 DDR4-2666 DDR4-2933 / 3200
Parameter Symbol Min Max Min Max Min Max Unit
Single-ended output SRQse 4 9 4 9 4 9 V/ns
slew rate
Notes: 1. SR = slew rate; Q = query output; se = single-ended signals.

2. In two cases a maximum slew rate of 12V/ns applies for a single DQ signal within a byte
lane:
• Case 1 is defined for a single DQ signal within a byte lane that is switching into a cer-
tain direction (either from HIGH-to-LOW or LOW-to-HIGH) while all remaining DQ sig-
nals in the same byte lane are static (they stay at either HIGH or LOW).
• Case 2 is defined for a single DQ signal within a byte lane that is switching into a cer-
tain direction (either from HIGH-to-LOW or LOW-to-HIGH) while all remaining DQ sig-
nals in the same byte lane are switching into the opposite direction (from LOW-to-
HIGH or HIGH-to-LOW, respectively). For the remaining DQ signal switching into the
opposite direction, the standard maximum limit of 9 V/ns applies.
Differential Outputs
Table 115: Differential Output Levels

AC differential output high measurement level (for output slew VOH,diff(AC) 0.3 × VDDQ V
rate)
AC differential output low measurement level (for output slew VOL,diff(AC) –0.3 × VDDQ V
rate)
Note: 1. The swing of ±0.3 × VDDQ is based on approximately 50% of the static single-ended out-
put peak-to-peak swing with a driver impedance of RZQ/7 and an effective test load of
50˖ to VTT = VDDQ at each differential output.
Using the same reference load used for timing measurements, output slew rate for fall-
ing and rising edges is defined and measured between V OL,diff(AC) and V OH,diff(AC) for dif-
ferential signals.
Table 116: Differential Output Slew Rate Definition
Measured
Differential output slew rate for rising edge VOL,diff(AC) VOH,diff(AC) [VOH,diff(AC) - VOL,diff(AC)@˂TRdiff
Differential output slew rate for falling edge VOH,diff(AC) VOL,diff(AC) [VOH,diff(AC) - VOL,diff(AC)@˂TFdiff
Levels
Figure 228: Differential Output Slew Rate Definition

TRdiff
Differential Input Voltage (DQS_t, DQS_c) VOH,diff(AC)
VOL,diff(AC)
TFdiff
Table 117: Differential Output Slew Rate

For RON = RZQ/7
DDR4-1600 / 1866 /
2133 / 2400 DDR4-2666 DDR4-2933 / 3200
Differential output slew SRQdiff 8 18 8 18 8 18 V/ns
rate
Note: 1. SR = slew rate; Q = query output; diff = differential signals.
Reference Load for AC Timing and Output Slew Rate

The effective reference load of 50˖ to V TT = V DDQ and driver impedance of RZQ/7 for
each output was used in defining the relevant AC timing parameters of the device as
well as output slew rate measurements.
RON nominal of DQ, DQS_t and DQS_c drivers uses 34 ohms to specify the relevant AC
timing parameter values of the device. The maximum DC high level of output signal =
1.0 × V DDQ, the minimum DC low level of output signal = { 34 /( 34 + 50 ) } × V DDQ = 0.4 ×
VDDQ.
The nominal reference level of an output signal can be approximated by the following:
The center of maximum DC high and minimum DC low = { ( 1 + 0.4 ) / 2 } × V DDQ = 0.7 ×
VDDQ. The actual reference level of output signal might vary with driver R ON and refer-
ence load tolerances. Thus, the actual reference level or midpoint of an output signal is
at the widest part of the output signal’s eye.
Levels
Figure 229: Reference Load For AC Timing and Output Slew Rate
VDDQ VTT = VDDQ
DQ, DQS_t, DQS_c,

DM, TDQS_t, TDQS_c
CK_t, CK_c
DUT
RTT = 50ȍ
VSSQ
Timing reference point
Connectivity Test Mode Output Levels
Table 118: Connectivity Test Mode Output Levels

DC output high measurement level (for IV curve linearity) VOH(DC) 1.1 × VDDQ V
DC output mid measurement level (for IV curve linearity) VOM(DC) 0.8 × VDDQ V
DC output low measurement level (for IV curve linearity) VOL(DC) 0.5 × VDDQ V
DC output below measurement level (for IV curve linearity) VOB(DC) 0.2 × VDDQ V
AC output high measurement level (for output slew rate) VOH(AC) VTT + (0.1 × VDDQ) V
AC output low measurement level (for output slew rate) VOL(AC) VTT - (0.1 × VDDQ) V
Note: 1. Driver impedance of RZQ/7 and an effective test load of 50˖ to VTT = VDDQ.
Figure 230: Connectivity Test Mode Reference Test Load
VDDQ
DQ, DQS_t, DQS_c,
LDQS_t, LDQS_c, UDQS_t, UDQS_c,
DM, LDM, HDM, TDQS_t, TDQS_c
CT_Inputs
DUT 0.5 × VDDQ
RTT = 50 ȍ
VSSQ
Electrical Characteristics – AC and DC Output Driver Charac-
teristics
Figure 231: Connectivity Test Mode Output Slew Rate Definition
VOH(AC)
VTT
0.5 x VDD
VOL(AC)
TFoutput_CT TRoutput_CT
Table 119: Connectivity Test Mode Output Slew Rate
DDR4-1600 / 1866 / DDR4-2933 /

2133 / 2400 DDR4-2666 3200
Output signal falling time TF_output_CT – 10 – 10 – 10 ns/V
Output signal rising time TR_output_CT – 10 – 10 – 10 ns/V
Electrical Characteristics – AC and DC Output Driver Characteristics

Connectivity Test Mode Output Driver Electrical Characteristics
The DDR4 driver supports special values during connectivity test mode. These R ON val-
ues are referenced in this section. A functional representation of the output buffer is
shown in the figure below.
teristics
Figure 232: Output Driver During Connectivity Test Mode
Chip in drive mode

Output driver
VDDQ
IPU_CT
To
RONPU_CT
other
circuitry
DQ
like
RCV, IOUT
RONPD_CT
...
VOUT
IPD_CT
VSSQ
The output driver impedance, RON, is determined by the value of the external reference
resistor RZQ as follows: RON = RZQ/7. This targets 34˖ with nominal RZQ ˖; however,
connectivity test mode uses uncalibrated drivers and only a maximum target is defined.
Mismatch between pull up and pull down is undefined.
The individual pull-up and pull-down resistors (RONPu_CT and RONPd_CT) are defined as
follows:
RONPu_CT when RONPd_CT is off:
9''49287
52138B&7
,287
RONPD_CT when RONPU_CT is off:

9287
5213'B&7
,287
teristics
Table 120: Output Driver Electrical Characteristics During Connectivity Test Mode
Assumes RZQ ˖; ZQ calibration not required
RON,nom_CT Resistor VOUT Min Nom Max Unit
VOB(DC) = 0.2 × VDDQ N/A N/A 1.9 RZQ/7
VOL(DC) = 0.5 × VDDQ N/A N/A 2.0 RZQ/7
RONPD_CT
VOM(DC) = 0.8 × VDDQ N/A N/A 2.2 RZQ/7
VOH(DC) = 1.1 × VDDQ N/A N/A 2.5 RZQ/7
˖
VOB(DC) = 0.2 × VDDQ N/A N/A 1.9 RZQ/7
VOL(DC) = 0.5 × VDDQ N/A N/A 2.0 RZQ/7
RONPU_CT
VOM(DC) = 0.8 × VDDQ N/A N/A 2.2 RZQ/7
VOH(DC) = 1.1 × VDDQ N/A N/A 2.5 RZQ/7
Output Driver Electrical Characteristics

The DDR4 driver supports two RON values. These R ON values are referred to as strong
mode (low RON˖) and weak mode (high RON˖). A functional representation of
the output buffer is shown in the figure below.
Figure 233: Output Driver: Definition of Voltages and Currents
Chip in drive mode

Output driver
VDDQ
IPU
To
RONPU
other
circuitry
DQ
like
RCV, IOUT
RONPD
...
VOUT
IPD
VSSQ
The output driver impedance, RON, is determined by the value of the external reference
resistor RZQ as follows: RON(34) = RZQ/7, or RON(48) = RZQ/5. This provides either a nomi-
nal 34.3˖±10% or 48˖±10% with nominal RZQ ˖
The individual pull-up and pull-down resistors (RONPu and RONPd) are defined as fol-
lows:
RONPu when RONPd is off:
teristics
VDDQ - VOUT
RONPU =
IOUT
RONPD when RONPU is off:
VOUT
RONPD =
IOUT
˖) Output Driver Electrical Characteristics

Table 121: Strong Mode (34˖
Assumes RZQ ˖; Entire operating temperature range after proper ZQ calibration
RON,nom Resistor VOUT Min Nom Max Unit Notes
VOL(DC) = 0.5 × VDDQ 0.73 1.00 1.10 RZQ/7 1, 2, 3
RON34PD VOM(DC) = 0.8 × VDDQ 0.83 1.00 1.10 RZQ/7 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.83 1.00 1.25 RZQ/7 1, 2, 3
˖
VOL(DC) = 0.5 × VDDQ 0.90 1.00 1.25 RZQ/7 1, 2, 3
RON34PU VOM(DC) = 0.8 × VDDQ 0.90 1.00 1.10 RZQ/7 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.80 1.00 1.10 RZQ/7 1, 2, 3
Mismatch between pull-up and pull- VOM(DC) = 0.8 × VDDQ 10 – 23 % 1, 2, 3, 4,
down, MMPUPD 6, 7
Mismatch between DQ to DQ within VOM(DC) = 0.8 × VDDQ – – 10 % 1, 2, 3, 4,
byte variation pull-up, MMPUdd 5
Mismatch between DQ to DQ within VOM(DC) = 0.8 × VDDQ - – 10 % 1, 2, 3, 4,
byte variation pull-down, MMPDdd 6, 7
Notes: 1. The tolerance limits are specified after calibration with stable voltage and temperature.
For the behavior of the tolerance limits if temperature or voltage changes after calibra-
tion, see following section on voltage and temperature sensitivity.
2. The tolerance limits are specified under the condition that VDDQ = VDD and that VSSQ =
VSS.
3. Micron recommends calibrating pull-down and pull-up output driver impedances at 0.8
× VDDQ. Other calibration schemes may be used to achieve the linearity specification
shown above; for example, calibration at 0.5 × VDDQ and 1.1 VDDQ.
4. DQ-to-DQ mismatch within byte variation for a given component including DQS_t and
DQS_c (characterized).
5. Measurement definition for mismatch between pull-up and pull-down, MMPUPD:
Measure both RONPU and RONPD at 0.8 × VDDQ separately; RON,nom is the nominal RON val-
ue:
RONPU - RONPD
MMPUPD = × 100
RON,nom
6. RON variance range ratio to RON nominal value in a given component, including DQS_t
and DQS_c:
teristics
RONPU,max - RONPU,min
MMPUDD = × 100
RON,nom
RONPD,max - RONPD,min
MMPDDD = × 100
RON,nom
7. The lower and upper bytes of a x16 are each treated on a per byte basis.
8. The minimum values are derated by 9% when the device operates between –40°C and
0°C (TC).
˖) Output Driver Electrical Characteristics

Table 122: Weak Mode (48˖
Assumes RZQ ˖; Entire operating temperature range after proper ZQ calibration
RON,nom Resistor VOUT Min Nom Max Unit Notes
˖ RON48PD VOL(DC) = 0.5 × VDDQ 0.73 1.00 1.10 RZQ/5 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.83 1.00 1.10 RZQ/5 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.83 1.00 1.25 RZQ/5 1, 2, 3
RON48PU VOL(DC) = 0.5 × VDDQ 0.90 1.00 1.25 RZQ/5 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.90 1.00 1.10 RZQ/5 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.80 1.00 1.10 RZQ/5 1, 2, 3
Mismatch between pull-up and VOM(DC) = 0.8 × VDDQ 10 – 23 % 1, 2, 3, 4,
pull-down, MMPUPD 6, 7
Mismatch between DQ to DQ VOM(DC) = 0.8 × VDDQ – – 10 % 1, 2, 3, 4, 5
within byte variation pull-up,
MMPUdd
Mismatch between DQ to DQ VOM(DC) = 0.8 × VDDQ – – 10 % 1, 2, 3, 4,
within byte variation pull-down, 6, 7
MMPDdd
Notes: 1. The tolerance limits are specified after calibration with stable voltage and temperature.
For the behavior of the tolerance limits if temperature or voltage changes after calibra-
tion, see following section on voltage and temperature sensitivity.
2. The tolerance limits are specified under the condition that VDDQ = VDD and that VSSQ =
VSS.
3. Micron recommends calibrating pull-down and pull-up output driver impedances at 0.8
× VDDQ. Other calibration schemes may be used to achieve the linearity specification
shown above; for example, calibration at 0.5 × VDDQ and 1.1 VDDQ.
4. DQ-to-DQ mismatch within byte variation for a given component including DQS_t and
DQS_c (characterized).
5. Measurement definition for mismatch between pull-up and pull-down, MMPUPD:
Measure both RONPU and RONPD at 0.8 × VDDQ separately; RON,nom is the nominal RON val-
ue:
RONPU - RONPD
MMPUPD = × 100
RON,nom
6. RON variance range ratio to RON nominal value in a given component, including DQS_t
and DQS_c:
teristics
RONPU,max - RONPU,min
MMPUDD = × 100
RON,nom
RONPD,max - RONPD,min
MMPDDD = × 100
RON,nom
7. The lower and upper bytes of a x16 are each treated on a per byte basis.
8. The minimum values are derated by 9% when the device operates between –40°C and
0°C (TC).
Output Driver Temperature and Voltage Sensitivity

If temperature and/or voltage change after calibration, the tolerance limits widen ac-
cording to the equations and tables below.
˂T = T - T(@calibration); ˂V = V DDQ - V DDQ(@ calibration); V DD = V DDQ
Table 123: Output Driver Sensitivity Definitions
Symbol Min Max Unit

RONPU@ VOH(DC) 0.6 - dRONdTH × |˂T| - dRONdVH × |˂V| 1.1 _ dRONdTH × |˂T| + dRONdVH × |˂V| RZQ/6
RON@ VOM(DC) 0.9 - dRONdTM × |˂T| - dRONdVM × |˂V| 1.1 + dRONdTM × |˂T| + dRONdVM × |˂V| RZQ/6
RONPD@ VOL(DC) 0.6 - dRONdTL × |˂T| - dRONdVL × |˂V| 1.1 + dRONdTL × |˂T| + dRONdVL × |˂V| RZQ/6
Table 124: Output Driver Voltage and Temperature Sensitivity
Voltage and Temperature Range

Symbol Min Max Unit
dRONdTM 0 1.5 %/°C
dRONdVM 0 0.15 %/mV
dRONdTL 0 1.5 %/°C
dRONdVL 0 0.15 %/mV
dRONdTH 0 1.5 %/°C
dRONdVM 0 0.15 %/mV
Alert Driver
A functional representation of the alert output buffer is shown in the figure below. Out-
put driver impedance, RON, is defined as follows.
Electrical Characteristics – On-Die Termination Characteristics
Figure 234: Alert Driver

Alert driver
'5$0
Alert
IOUT
RONPD VOUT
IPD
VSSQ
RONPD when RONPU is off:
VOUT
RONPD =
IOUT
Table 125: Alert Driver Voltage
RON,nom Register VOUT Min Nom Max Unit

N/A RONPD VOL(DC) = 0.1 × VDDQ 0.3 N/A 1.2 RZQ/7
VOM(DC) = 0.8 × VDDQ 0.4 N/A 1.2 RZQ/7
VOH(DC) = 1.1 × VDDQ 0.4 N/A 1.4 RZQ/7
Note: 1. VDDQ voltage is at VDDQ(DC).

ODT Levels and I-V Characteristics
On-die termination (ODT) effective resistance settings are defined and can be selected
by any or all of the following options:
• MR1[10:8] (RTT(NOM)): Disable, 240 ohms, 120 ohms, 80 ohms, 60 ohms, 48 ohms, 40
ohms, and 34 ohms.
• MR2[11:9] (RTT(WR)): Disable, 240 ohms,120 ohms, and 80 ohms.
• MR5[8:6] (RTT(Park)): Disable, 240 ohms, 120 ohms, 80 ohms, 60 ohms, 48 ohms, 40
ohms, and 34 ohms.
ODT is applied to the following inputs:
• x4: DQ, DM_n, DQS_t, and DQS_c inputs.
• x8: DQ, DM_n, DQS_t, DQS_c, TDQS_t, and TDQS_c inputs.
• x16: DQ, LDM_n, UDM_n, LDQS_t, LDQS_c, UDQS_t, and UDQS_c inputs.
A functional representation of ODT is shown in the figure below.
Figure 235: ODT Definition of Voltages and Currents

Chip in termination mode
ODT
VDDQ
RTT
To other
circuitry
like RCV,
DQ
...
IOUT
VOUT
VSSQ
Table 126: ODT DC Characteristics
RTT VOUT Min Nom Max Unit Notes

240 ohm VOL(DC) = 0.5 × VDDQ 0.9 1 1.25 RZQ 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ 1, 2, 3
120 ohm VOL(DC) = 0.5 × VDDQ 0.9 1 1.25 RZQ/2 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/2 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/2 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/3 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/3 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/4 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/4 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/5 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/5 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/6 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/6 1, 2, 3
VOM(DC) = 0.8 × VDDQ 0.9 1 1.1 RZQ/7 1, 2, 3
VOH(DC) = 1.1 × VDDQ 0.8 1 1.1 RZQ/7 1, 2, 3
DQ-to-DQ mismatch VOM(DC) = 0.8 × VDDQ 0 – 10 % 1, 2, 4, 5, 6
within byte
Notes: 1. The tolerance limits are specified after calibration to 240 ohm ±1% resistor with stable
voltage and temperature. For the behavior of the tolerance limits if temperature or
voltage changes after calibration, see ODT Temperature and Voltage Sensitivity.
2. Micron recommends calibrating pull-up ODT resistors at 0.8 × VDDQ. Other calibration
schemes may be used to achieve the linearity specification shown here.
3. The tolerance limits are specified under the condition that VDDQ = VDD and VSSQ = VSS.
4. The DQ-to-DQ mismatch within byte variation for a given component including DQS_t
and DQS_c.
5. RTT variance range ratio to RTT nominal value in a given component, including DQS_t
and DQS_c.
RTT(MAX) - RTT(MIN)
DQ-to-DQ mismatch = × 100
RTT(NOM)
6. DQ-to-DQ mismatch for a x16 device is treated as two separate bytes.
7. For IT, AT, and UT devices, the minimum values are derated by 9% when the device op-
erates between –40°C and 0°C (TC).
ODT Temperature and Voltage Sensitivity

If temperature and/or voltage change after calibration, the tolerance limits widen ac-
cording to the following equations and tables.
˂T = T - T(@ calibration); ˂V = V DDQ - V DDQ(@ calibration); V DD = V DDQ
Table 127: ODT Sensitivity Definitions
Parameter Min Max Unit

RTT@ 0.9 - dRTTdT × |˂T| - dRTTdV × |˂V| 1.6 + dRTTdTH × |˂T| + dRTTdVH × |˂V| RZQ/n
Table 128: ODT Voltage and Temperature Sensitivity
Parameter Min Max Unit

dRTTdT 0 1.5 %/°C
dRTTdV 0 0.15 %/mV
ODT Timing Definitions

The reference load for ODT timings is different than the reference load used for timing
measurements.
Figure 236: ODT Timing Reference Load

VDDQ
DQ, DQS_t, DQS_c,

DM, TDQS_t, TDQS_c
CK_t, CK_c
DUT
RTT = 50ȍ
VSSQ VTT = VSSQ
ODT Timing Definitions

Definitions for tADC, tAONAS, and tAOFAS are provided in the Table 129 (page 304) and
shown in Figure 237 (page 305) and Figure 239 (page 306). Measurement reference set-
tings are provided in the subsequent Table 130 (page 304).
The tADC for the dynamic ODT case and read disable ODT cases are represented by
tADC of Direct ODT Control case.
Table 129: ODT Timing Definitions
Parameter Begin Point Definition End Point Definition Figure

tADC Rising edge of CK_t, CK_c defined by the end point of Extrapolated point at VRTT,nom Figure 237
DODTLoff (page 305)
Rising edge of CK_t, CK_c defined by the end point of Extrapolated point at VSSQ Figure 237
DODTLon (page 305)
Rising edge of CK_t, CK_c defined by the end point of Extrapolated point at VRTT,nom Figure 238
ODTLcnw (page 305)
Rising edge of CK_t, CK_c defined by the end point of Extrapolated point at VSSQ Figure 238
ODTLcwn4 or ODTLcwn8 (page 305)
tAONAS Rising edge of CK_t, CK_c with ODT being first registered Extrapolated point at VSSQ Figure 239
HIGH (page 306)
tAOFAS Rising edge of CK_t, CK_c with ODT being first registered Extrapolated point at VRTT,nom Figure 239
LOW (page 306)
Table 130: Reference Settings for ODT Timing Measurements
Measure
Parameter RTT(Park) RTT(NOM) RTT(WR) VSW1 VSW2 Note
tADC Disable RZQ˖ – 0.20V 0.40V 1, 2, 4
– RZQ˖ High-Z 0.20V 0.40V 1, 3, 5
tAONAS Disable RZQ˖ – 0.20V 0.40V 1, 2, 6
tAOFAS Disable RZQ˖ – 0.20V 0.40V 1, 2, 6
Notes: 1. MR settings are as follows: MR1 has A10 = 1, A9 = 1, A8 = 1 for RTT(NOM) setting; MR5 has
A8 = 0, A7 = 0, A6 = 0 for RTT(Park) setting; and MR2 has A11 = 0, A10 = 1, A9 = 1 for
RTT(WR) setting.
2. ODT state change is controlled by ODT pin.
3. ODT state change is controlled by a WRITE command.
4. Refer to Figure 237 (page 305).
Figure 237: tADC Definition with Direct ODT Control
DODTLoff Begin point: Rising edge DODTLon Begin point: Rising edge
of CK_t, CK_c defined of CK_t, CK_c defined
by the end point of by the end point of
DODTLoff DODTLon
CK_c
CK_t
tADC tADC
End point: Extrapolated

VRTT,nom point at VRTT,nom VRTT,nom
Vsw2
DQ, DM
DQS_t, DQS_c Vsw1
TDQS_t, TDQS_c End point: Extrapolated
VSSQ VSSQ point at VSSQ
Figure 238: tADC Definition with Dynamic ODT Control
ODTLcnw Begin point: Rising edge ODTLcnw4/8 Begin point: Rising edge
of CK_t, CK_c defined of CK_t, CK_c defined
by the end point of by the end point of
ODTLcnw ODTLcnw4 or ODTLcnw8
CK_c
CK_t
tADC tADC

VRTT,nom point at VRTT,nom VRTT,nom
Vsw2
DQ, DM
DQS_t, DQS_c Vsw1
Figure 239: tAOFAS and tAONAS Definitions

Rising edge of CK_t, CK_c Rising edge of CK_t, CK_c
with ODT being first with ODT being first
registered LOW registered HIGH
CK_c
CK_t
tAOFAS tAONAS

VRTT,nom point at VRTT_NOM VRTT,nom
Vsw2
DQ, DM
DQS_t, DQS_c Vsw1
DRAM Package Electrical Specifications
Table 131: DRAM Package Electrical Specifications for x4 and x8 Devices
1600/1866/2133/
2400/2666 2933 3200
Input/ Zpkg ZIO 45 85 48 85 48 85 ohm 1, 2, 4
output Package delay TdIO 14 42 14 40 14 40 ps 1, 3, 4
Lpkg LIO – 3.3 – 3.3 – 3.3 nH 10
Cpkg CIO – 0.78 – 0.78 – 0.78 pF 11
DQS_t, Zpkg ZIO DQS 45 85 48 85 48 85 ohm 1, 2
DQS_c Package delay TdIO DQS 14 42 14 40 14 40 ps 1, 3
Delta Zpkg DZIO DQS – 10 – 10 – 10 ohm 1, 2, 6
Delta delay DTdIO DQS – 5 – 5 – 5 ps 1, 3, 6
Lpkg LIO DQS – 3.3 – 3.3 – 3.3 nH 10
Cpkg CIO DQS – 0.78 – 0.78 – 0.78 pF 11
Input CTRL Zpkg ZI CTRL 50 90 50 90 50 90 ohm 1, 2, 8
pins Package delay TdI CTRL 14 42 14 40 14 40 ps 1, 3, 8
Lpkg LI CTRL – 3.4 – 3.4 – 3.4 nH 10
Cpkg CI CTRL – 0.7 – 0.7 – 0.7 pF 11
Input CMD Zpkg ZI ADD CMD 50 90 50 90 50 90 ohm 1, 2, 7
ADD pins Package delay TdI ADD CMD 14 45 14 40 14 40 ps 1, 3, 7
Lpkg LI ADD CMD – 3.6 – 3.6 – 3.6 nH 10
Cpkg CI ADD CMD – 0.74 – 0.74 – 0.74 pF 11
CK_t, CK_c Zpkg ZCK 50 90 50 90 50 90 ohm 1, 2
Package delay TdCK 14 42 14 42 14 42 ps 1, 3
Delta Zpkg DZDCK – 10 – 10 – 10 ohm 1, 2, 5
Delta delay DTdDCK – 5 – 5 – 5 ps 1, 3, 5
Lpkg LI CLK – 3.4 – 3.4 – 3.4 nH 10
Cpkg CI CLK – 0.7 – 0.7 – 0.7 pF 11
ZQ Zpkg ZO ZQ – 100 – 100 – 100 ohm 1, 2
ZQ delay TdO ZQ 20 90 20 90 20 90 ps 1, 3
ALERT Zpkg ZO ALERT 40 100 40 100 40 100 ohm 1, 2
ALERT delay TdO ALERT 20 55 20 55 20 55 ps 1, 3
Notes: 1. The package parasitic (L and C) are validated using package only samples. The capaci-
tance is measured with VDD, VDDQ, VSS, and VSSQ shorted with all other signal pins float-
ing. The inductance is measured with VDD, VDDQ, VSS, and VSSQ shorted and all other sig-
nal pins shorted at the die, not pin, side.
2. Package-only impedance (Zpkg) is calculated based on the Lpkg and Cpkg total for a
given pin where: Zpkg (total per pin) = SQRT (Lpkg/Cpkg).
3. Package-only delay (Tpkg) is calculated based on Lpkg and Cpkg total for a given pin
where: Tdpkg (total per pin) = SQRT (Lpkg × Cpkg).
4. ZIO and TdIO apply to DQ, DM, TDQS_t and TDQS_c.
5. Absolute value of ZCK_t, ZCK_c for impedance (Z) or absolute value of TdCK_t, TdCK_c
for delay (Td).
6. Absolute value of ZIO (DQS_t), ZIO (DQS_c) for impedance (Z) or absolute value of TdIO
(DQS_t), TdIO (DQS_c) for delay (Td).
7. ZI ADD CMD and TdI ADD CMD apply to A[17:0], BA[1:0], BG[1:0], RAS_n CAS_n, and WE_n.
8. ZI CTRL and TdI CTRL apply to ODT, CS_n, and CKE.
9. Package implementations will meet specification if the Zpkg and package delay fall
within the ranges shown, and the maximum Lpkg and Cpkg do not exceed the maxi-
mum values shown.
10. It is assumed that Lpkg can be approximated as Lpkg = ZO × Td.
11. It is assumed that Cpkg can be approximated as Cpkg = Td/ZO.
Table 132: DRAM Package Electrical Specifications for x16 Devices
1600/1866/2133/
2400/2666 2933 3200
Input/ Zpkg ZIO 45 85 45 85 45 85 ohm 1, 2, 4
output Package delay TdIO 14 45 14 45 14 45 ps 1, 3, 4
Lpkg LIO – 3.4 – 3.4 – 3.4 nH 11
Cpkg CIO – 0.82 – 0.82 – 0.82 pF 11
LDQS_t/ Zpkg ZIO DQS 45 85 45 85 45 85 ohm 1, 2
LDQS_c/ Package delay TdIO DQS 14 45 14 45 14 45 ps 1, 3
UDQS_t/
Lpkg LIO DQS – 3.4 – 3.4 – 3.4 nH 11
UDQS_c
Cpkg CIO DQS – 0.82 – 0.82 – 0.82 pF 11
LDQS_t/ Delta Zpkg DZIO DQS – 10.5 – 10.5 – 10.5 ohm 1, 2, 6
LDQS_c, Delta delay DTdIO DQS – 5 – 5 – 5 ps 1, 3, 6
UDQS_t/
UDQS_c,
Input CTRL Zpkg ZI CTRL 50 90 50 90 50 90 ohm 1, 2, 8
pins Package delay TdI CTRL 14 42 14 42 14 42 ps 1, 3, 8
Lpkg LI CTRL – 3.4 – 3.4 – 3.4 nH 11
Cpkg CI CTRL – 0.7 – 0.7 – 0.7 pF 11
Input CMD Zpkg ZI ADD CMD 50 90 50 90 50 90 ohm 1, 2, 7
ADD pins Package delay TdI ADD CMD 14 52 14 52 14 52 ps 1, 3, 7
Lpkg LI ADD CMD – 3.9 – 3.9 – 3.9 nH 11
Cpkg CI ADD CMD – 0.86 – 0.86 – 0.86 pF 11
CK_t, CK_c Zpkg ZCK 50 90 50 90 50 90 ohm 1, 2
Package delay TdCK 14 42 14 42 14 42 ps 1, 3
Delta Zpkg DZDCK – 10.5 – 10.5 – 10.5 ohm 1, 2, 5
Delta delay DTdDCK – 5 – 5 – 5 ps 1, 3, 5
Input CLK Lpkg LI CLK – 3.4 – 3.4 – 3.4 nH 11
Cpkg CI CLK – 0.7 – 0.7 – 0.7 pF 11
ZQ Zpkg ZO ZQ – 100 – 100 – 100 ohm 1, 2
ZQ delay TdO ZQ 20 90 20 90 20 90 ps 1, 3
Table 132: DRAM Package Electrical Specifications for x16 Devices (Continued)
1600/1866/2133/
2400/2666 2933 3200
ALERT Zpkg ZO ALERT 40 100 40 100 40 100 ohm 1, 2
ALERT delay TdO ALERT 20 55 20 55 20 55 ps 1, 3
Notes: 1. The package parasitic (L and C) are validated using package only samples. The capaci-
tance is measured with VDD, VDDQ, VSS, and VSSQ shorted with all other signal pins float-
ing. The inductance is measured with VDD, VDDQ, VSS, and VSSQ shorted and all other sig-
nal pins shorted at the die, not pin, side.
2. Package-only impedance (Zpkg) is calculated based on the Lpkg and Cpkg total for a
given pin where: Zpkg (total per pin) = SQRT (Lpkg/Cpkg).
3. Package-only delay (Tpkg) is calculated based on Lpkg and Cpkg total for a given pin
where: Tdpkg (total per pin) = SQRT (Lpkg × Cpkg).
4. ZIO and TdIO apply to DQ, DM, TDQS_t and TDQS_c.
5. Absolute value of ZCK_t, ZCK_c for impedance (Z) or absolute value of TdCK_t, TdCK_c
for delay (Td).
6. Absolute value of ZIO (DQS_t), ZIO (DQS_c) for impedance (Z) or absolute value of TdIO
(DQS_t), TdIO (DQS_c) for delay (Td).
7. ZI ADD CMD and TdI ADD CMD apply to A[17:0], BA[1:0], BG[1:0], RAS_n CAS_n, and WE_n.
8. ZI CTRL and TdI CTRL apply to ODT, CS_n, and CKE.
9. Package implementations will meet specification if the Zpkg and package delay fall
within the ranges shown, and the maximum Lpkg and Cpkg do not exceed the maxi-
mum values shown.
10. It is assumed that Lpkg can be approximated as Lpkg = ZO × Td.
11. It is assumed that Cpkg can be approximated as Cpkg = Td/ZO.
Table 133: Pad Input/Output Capacitance
DDR4-1600, DDR4-2400,
1866, 2133 2666 DDR4-2933 DDR4-3200
Parameter Symbol Min Max Min Max Min Max Min Max Unit Notes
Input/output capacitance: CIO 0.55 1.4 0.55 1.15 0.55 1.00 0.55 1.00 pF 1, 2, 3
DQ, DM, DQS_t, DQS_c,
TDQS_t, TDQS_c
Input capacitance: CK_t and CCK 0.2 0.8 0.2 0.7 0.2 0.7 0.2 0.7 pF 2, 3
CK_c
Input capacitance delta: CK_t CDCK 0 0.05 0 0.05 0 0.05 0 0.05 pF 2, 3, 6
and CK_c
Input/output capacitance del- CDDQS 0 0.05 0 0.05 0 0.05 0 0.05 pF 2, 3, 5
ta: DQS_t and DQS_c
Input capacitance: CTRL, CI 0.2 0.8 0.2 0.7 0.2 0.6 0.2 0.55 pF 2, 3, 4
ADD, CMD input-only pins
Input capacitance delta: All CDI_CTRL –0.1 0.1 –0.1 0.1 –0.1 0.1 –0.1 0.1 pF 2, 3, 8,
CTRL input-only pins 9
Input capacitance delta: All CDI_ADD_CM –0.1 0.1 –0.1 0.1 –0.1 0.1 –0.1 0.1 pF 1, 2, 10,
ADD/CMD input-only pins D 11
Input/output capacitance del- CDIO –0.1 0.1 –0.1 0.1 –0.1 0.1 –0.1 0.1 pF 1, 2, 3,
ta: DQ, DM, DQS_t, DQS_c, 4
TDQS_t, TDQS_c
Input/output capacitance: CALERT 0.5 1.5 0.5 1.5 0.5 1.5 0.5 1.5 pF 2, 3
ALERT pin
Input/output capacitance: ZQ CZQ – 2.3 – 2.3 – 2.3 – 2.3 pF 2, 3, 12
pin
Input/output capacitance: CTEN 0.2 2.3 0.2 2.3 0.2 2.3 0.2 2.3 pF 2, 3, 13
TEN pin
Notes: 1. Although the DM, TDQS_t, and TDQS_c pins have different functions, the loading
matches DQ and DQS.
2. This parameter is not subject to a production test; it is verified by design and characteri-
zation. The capacitance is measured according to the JEP147 specification, “Procedure
for Measuring Input Capacitance Using a Vector Network Analyzer (VNA),” with VDD,
VDDQ, VSS, and VSSQ applied and all other pins floating (except the pin under test, CKE,
RESET_n and ODT, as necessary). VDD = VDDQ = 1.2V, VBIAS = VDD/2 and on-die termination
off. Measured data is rounded using industry standard half-rounded up methodology to
the nearest hundredth of the MSB.
3. This parameter applies to monolithic die, obtained by de-embedding the package L and
C parasitics.
4. CDIO = CIO(DQ, DM) - 0.5 × (CIO(DQS_t) + CIO(DQS_c)).
5. Absolute value of CIO (DQS_t), CIO (DQS_c)
6. Absolute value of CCK_t, CCK_c
7. CI applies to ODT, CS_n, CKE, A[17:0], BA[1:0], BG[1:0], RAS_n, CAS_n, ACT_n, PAR and
WE_n.
8. CDI_CTRL applies to ODT, CS_n, and CKE.
9. CDI_CTRL = CI(CTRL) - 0.5 × (CI(CLK_t) + CI(CLK_c)).
Thermal Characteristics
10. CDI_ADD_CMD applies to A[17:0], BA1:0], BG[1:0], RAS_n, CAS_n, ACT_n, PAR and WE_n.
11. CDI_ADD_CMD = CI(ADD_CMD) - 0.5 × (CI(CLK_t) + CI(CLK_c)).
12. Maximum external load capacitance on ZQ pin: 5pF.
13. Only applicable if TEN pin does not have an internal pull-up.
Thermal Characteristics
Table 134: Thermal Characteristics
Parameter/Condition Value Units Symbol Notes

Operating case temperature: 0 to +85 °C TC 1, 2, 3
Commercial 0 to +95 °C TC 1, 2, 3, 4
Operating case temperature: –40 to +95 °C TC 1, 2, 3, 4
Industrial
Automotive
Ultra-high
REV B 78-ball “RH” Junction-to-case (TOP) 7.7 °C/W ˆJC 5
Junction-to-board 20.9 °C/W ˆJB
96-ball “GE” Junction-to-case (TOP) 5.0 °C/W ˆJC 5
REV F 78-ball “SA” Junction-to-case (TOP) 4.9 °C/W ˆJC 5
96-ball “LY” Junction-to-case (TOP) 4.8 °C/W ˆJC 5
Notes: 1. MAX operating case temperature. TC is measured in the center of the package.
2. A thermal solution must be designed to ensure the DRAM device does not exceed the
maximum TC during operation.
3. Device functionality is not guaranteed if the DRAM device exceeds the maximum TC dur-
ing operation.
4. If TC exceeds 85°C, the DRAM must be refreshed externally at 2x refresh, which is a 3.9μs
interval refresh rate.
5. The thermal resistance data is based off of a number of samples from multiple lots and
should be viewed as a typical number.
Current Specifications – Measurement Conditions
Figure 240: Thermal Measurement Point
TC test point
(L/2)
(W/2)

IDD, IPP, and IDDQ Measurement Conditions
IDD, IPP, and IDDQ measurement conditions, such as test load and patterns, are defined
in this section.
• IDD currents (IDD0, IDD1, IDD2N, IDD2NT, IDD2P, IDD2Q, IDD3N, IDD3P, IDD4R, IDD4W, IDD5R,
IDD6N, IDD6E, IDD6R, IDD6A, IDD7, and IDD8) are measured as time-averaged currents
with all V DD balls of the device under test grouped together.
• IPP currents are IPP3N for standby cases (IDD2N, IDD2NT, IDD2P, IDD2Q, IDD3N, IDD3P, IDD8),
IPP0 for active cases (IDD0,IDD1, IDD4R, IDD4W), IPP5R for the distributed refresh case
(IDD5R), IPP6x for self refresh cases (IDD6N, IDD6E, IDD6R, IDD6A) and IPP7 for the operating
bank interleave read case (IDD7). These have the same definitions as the I DD currents
referenced but are measured on the V PP supply.
• IDDQ currents are measured as time-averaged currents with V DDQ balls of the device
under test grouped together. Micron does not specify IDDQ currents.
• IPP and IDDQ currents are not included in IDD currents, IDD and IDDQ currents are not
included in IPP currents, and IDD and IPP currents are not included in IDDQ currents.
Note: IDDQ values cannot be directly used to calculate the I/O power of the device. They
can be used to support correlation of simulated I/O power to actual I/O power. In
DRAM module application, IDDQ cannot be measured separately because V DD and V DDQ
are using a merged-power layer in the module PCB.
The following definitions apply for IDD, IPP and IDDQ measurements.
• “0” and “LOW” are defined as V IN ืVIL(AC)max
• “1” and “HIGH” are defined as V IN ุVIH(AC)min
• “Midlevel” is defined as inputs V REF = V DD/2
• Timings used for IDD, IPP and IDDQ measurement-loop patterns are provided in the
Current Test Definition and Patterns section.
• Basic IDD, IPP, and IDDQ measurement conditions are described in the Current Test
Definition and Patterns section.
• Detailed IDD, IPP, and IDDQ measurement-loop patterns are described in the Current
Test Definition and Patterns section.
• Current measurements are done after properly initializing the device. This includes,
but is not limited to, setting:
RON = RZQ/7 (34 ohm in MR1);
Qoff = 0B (output buffer enabled in MR1);
RTT(NOM) = RZQ/6 (40 ohm in MR1);
RTT(WR) = RZQ/2 (120 ohm in MR2);
RTT(Park) = disabled;
TDQS feature disabled in MR1; CRC disabled in MR2; CA parity feature disabled in
MR3; Gear-down mode disabled in MR3; Read/Write DBI disabled in MR5; DM disa-
bled in MR5
• Define D = {CS_n, RAS_n, CAS_n, WE_n}: = {HIGH, LOW, LOW, LOW}; apply BG/BA
changes when directed.
• Define D_n = {CS_n, RAS_n, CAS_n, WE_n}: = {HIGH, HIGH, HIGH, HIGH}; apply in-
vert of BG/BA changes when directed above.
Note: The measurement-loop patterns must be executed at least once before actual cur-
rent measurements can be taken.
Figure 241: Measurement Setup and Test Load for IDDx, IPPx, and IDDQx
IDD IPP IDDQ
VDD VPP VDDQ

RESET_n
CK_t/CK_c
DDR4
DQS_t, DQS_c
CKE SDRAM
CS_n DQ
C DM_n
ACT_n, RAS_n, CAS_n, WE_n
A, BG, BA
ODT
ZQ V VSSQ
SS
Figure 242: Correlation: Simulated Channel I/O Power to Actual Channel I/O Power
Applic ation-s pe c ific
memory c ha nne l I DD Q
env ironmen t tes t loa d
C hanne l I/O I DD Q IDD Q

pow er simulation simulation meas ure ment
C or relation
C orre c tion
C hanne l I/O
pow er n umber
Note: 1. Supported by IDDQ measurement.
IDD Definitions
Table 135: Basic IDD, IPP, and IDDQ Measurement Conditions
Symbol Description
IDD0 Operating One Bank Active-Precharge Current (AL = 0)
CKE: HIGH; External clock: On; tCK, nRC, nRAS, CL: see the previous table; BL: 8;1 AL: 0; CS_n: HIGH between
ACT and PRE; Command, address, bank group address, bank address inputs: partially toggling according to the
next table; Data I/O: VDDQ; DM_n: stable at 0; Bank activity: cycling with one bank active at a time: 0, 0, 1, 1, 2,
2, ... (see the IDD0 Measurement-Loop Pattern table); Output buffer and RTT: enabled in mode registers;2 ODT
signal: stable at 0; Pattern details: see the IDD0 Measurement-Loop Pattern table
IPP0 Operating One Bank Active-Precharge IPP Current (AL = 0)
Same conditions as IDD0 above
IDD1 Operating One Bank Active-Read-Precharge Current (AL = 0)
CKE: HIGH; External clock: on; tCK, nRC, nRAS, nRCD, CL: see the previous table; BL: 8;1, 5 AL: 0; CS_n: HIGH
between ACT, RD, and PRE; Command, address, bank group address, bank address inputs, Data I/O: partially
toggling according to the IDD1 Measurement-Loop Pattern table; DM_n: stable at 0; Bank activity: cycling with
one bank active at a time: 0, 0, 1, 1, 2, 2, ... (see the following table); Output buffer and RTT: enabled in mode
registers;2 ODT Signal: stable at 0; Pattern details: see the IDD1 Measurement-Loop Pattern table
IDD2N Precharge Standby Current (AL = 0)
CKE: HIGH; External clock: On; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n: stable at 1; Command, ad-
dress, bank group address, bank address Inputs: partially toggling according to the IDD2N and IDD3N Measure-
ment-Loop Pattern table; Data I/O: VDDQ; DM_n: stable at 1; Bank activity: all banks closed; Output buffer and
RTT: enabled in mode registers;2 ODT signal: stable at 0; Pattern details: see the IDD2N and IDD3N Measurement-
Loop Pattern table
Table 135: Basic IDD, IPP, and IDDQ Measurement Conditions (Continued)
Symbol Description
IDD2NT Precharge Standby ODT Current
CKE: HIGH; External clock: on; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n: stable at 1; Command, ad-
dress, bank gropup address, bank address inputs: partially toggling according to the IDD2NT Measurement-Loop
Pattern table; Data I/O: VSSQ; DM_n: stable at 1; Bank activity: all banks closed; Output buffer and RTT: enabled
in mode registers;2 ODT signal: toggling according to the IDD2NT Measurement-Loop Pattern table; Pattern de-
tails: see the IDD2NT Measurement-Loop Pattern table
IDD2P Precharge Power-Down Current
CKE: LOW; External clock: on; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n: stable at 1; Command, ad-
dress, bank group address, bank address inputs: stable at 0; Data I/O: VDDQ; DM_n: stable at 1; Bank activity: all
banks closed; Output buffer and RTT: Enabled in mode registers;2 ODT signal: stable at 0
IDD2Q Precharge Quiet Standby Current
banks closed; Output buffer and RTT: Enabled in mode registers;2 ODT signal: stable at 0
IDD3N Active Standby Current (AL = 0)
dress, bank group address, bank address inputs: partially toggling according to the IDD2N and IDD3N Measure-
ment-Loop Pattern table; Data I/O: VDDQ; DM_n: stable at 1; Bank activity: all banks open; Output buffer and
RTT: Enabled in mode registers;2 ODT signal: stable at 0; Pattern details: see the IDD2N and IDD3N Measurement-
Loop Pattern table
IPP3N Active Standby IPP3N Current (AL = 0)
Same conditions as IDD3N above
IDD3P Active Power-Down Current (AL = 0)
CKE: LOW; External clock: on; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n: stable at 1; Command, ad-
banks open; Output buffer and RTT: Enabled in mode registers;2 ODT signal: stable at 0
IDD4R Operating Burst Read Current (AL = 0)
CKE: HIGH; External clock: on; tCK, CL: see the previous table; BL: 8;15 AL: 0; CS_n: HIGH between RD; Com-
mand, address, bank group address, bank address inputs: partially toggling according to the IDD4R Measure-
ment-Loop Pattern table; Data I/O: seamless read data burst with different data between one burst and the
next one according to the IDD4R Measurement-Loop Pattern table; DM_n: stable at 1; Bank activity: all banks
open, RD commands cycling through banks: 0, 0, 1, 1, 2, 2, ... (see the IDD4R Measurement-Loop Pattern table);
Output buffer and RTT: Enabled in mode registers;2 ODT signal: stable at 0; Pattern details: see the IDD4R Meas-
urement-Loop Pattern table
IDD4W Operating Burst Write Current (AL = 0)
CKE: HIGH; External clock: on; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n: HIGH between WR; Com-
mand, address, bank group address, bank address inputs: partially toggling according to the IDD4W Measure-
ment-Loop Pattern table; Data I/O: seamless write data burst with different data between one burst and the
next one according to the IDD4W Measurement-Loop Pattern table; DM: stable at 0; Bank activity: all banks
open, WR commands cycling through banks: 0, 0, 1, 1, 2, 2, ... (see IDD4W Measurement-Loop Pattern table);
Output buffer and RTT: enabled in mode registers (see note2); ODT signal: stable at HIGH; Pattern details: see
the IDD4W Measurement-Loop Pattern table
Table 135: Basic IDD, IPP, and IDDQ Measurement Conditions (Continued)
Symbol Description
IDD5R Distributed Refresh Current (1X REF)
CKE: HIGH; External clock: on; tCK, CL, nREFI: see the previous table; BL: 8;1 AL: 0; CS_n: HIGH between REF;
Command, address, bank group address, bank address inputs: partially toggling according to the IDD5R Meas-
urement-Loop Pattern table; Data I/O: VDDQ; DM_n: stable at 1; Bank activity: REF command every nREFI (see
the IDD5R Measurement-Loop Pattern table); Output buffer and RTT: enabled in mode registers2; ODT signal:
stable at 0; Pattern details: see the IDD5R Measurement-Loop Pattern table
IPP5R Distributed Refresh Current (1X REF)
Same conditions as IDD5R above
IDD6N Self Refresh Current: Normal Temperature Range
TC: 0–85°C; Auto self refresh (ASR): disabled;3 Self refresh temperature range (SRT): normal;4 CKE: LOW; Exter-
nal clock: off; CK_t and CK_c: LOW; CL: see the table above; BL: 8;1 AL: 0; CS_n, command, address, bank group
address, bank address, data I/O: VDDQ; DM_n: stable at 1; Bank activity: SELF REFRESH operation; Output buffer
and RTT: enabled in mode registers;2 ODT signal: midlevel
IDD6E Self Refresh Current: Extended Temperature Range 4
TC: 0–95°C; Auto self refresh (ASR): disabled4; Self refresh temperature range (SRT): extended;4 CKE: LOW; Ex-
ternal clock: off; CK_t and CK_c: LOW; CL: see the previous table; BL: 8;1 AL: 0; CS_n, command, address, group
bank address, bank address, data I/O: VDDQ; DM_n: stable at 1; Bank activity: EXTENDED TEMPERATURE SELF
REFRESH operation; Output buffer and RTT: enabled in mode registers;2 ODT signal: midlevel
IPP6x Self Refresh IPP Current
Same conditions as IDD6E above
IDD6R Self Refresh Current: Reduced Temperature Range
TC: 0–45°C; Auto self refresh (ASR): disabled; Self refresh temperature range (SRT): reduced;4 CKE: LOW; Exter-
nal clock: off; CK_t and CK_c: LOW; CL: see the previous table; BL: 8;1 AL: 0; CS_n, command, address, bank
group address, bank address, data I/O: VDDQ; DM_n: stable at 1; Bank activity: EXTENDED TEMPERATURE SELF
REFRESH operation; Output buffer and RTT: enabled in mode registers;2 ODT signal: midlevel
IDD7 Operating Bank Interleave Read Current
CKE: HIGH; External clock: on; tCK, nRC, nRAS, nRCD, nRRD, nFAW, CL: see the previous table; BL: 8;15 AL: CL -
1; CS_n: HIGH between ACT and RDA; Command, address, group bank adress, bank address inputs: partially
toggling according to the IDD7 Measurement-Loop Pattern table; Data I/O: read data bursts with different data
between one burst and the next one according to the IDD7 Measurement-Loop Pattern table; DM: stable at 1;
Bank activity: two times interleaved cycling through banks (0, 1, ...7) with different addressing, see the IDD7
Measurement-Loop Pattern table; Output buffer and RTT: enabled in mode registers;2 ODT signal: stable at 0;
Pattern details: see the IDD7 Measurement-Loop Pattern table
IPP7 Operating Bank Interleave Read IPP Current
Same conditions as IDD7 above
IDD8 Maximum Power Down Current
Place DRAM in MPSM then CKE: HIGH; External clock: on; tCK, CL: see the previous table; BL: 8;1 AL: 0; CS_n:
stable at 1; Command, address, bank group address, bank address inputs: stable at 0; Data I/O: VDDQ; DM_n:
stable at 1; Bank activity: all banks closed; Output buffer and RTT: Enabled in mode registers;2 ODT signal: sta-
ble at 0
Notes: 1. Burst length: BL8 fixed by MRS: set MR0[1:0] 00.

2. Output buffer enable: set MR1[12] 0 (output buffer enabled); set MR1[2:1] 00 (RON =
RZQ/7); RTT(NOM) enable: set MR1[10:8] 011 (RZQ/6); RTT(WR) enable: set MR2[11:9] 001
(RZQ/2), and RTT(Park) enable: set MR5[8:6] 000 (disabled).
3. Auto self refresh (ASR): set MR2[6] 0 to disable or MR2[6] 1 to enable feature.
4. Self refresh temperature range (SRT): set MR2[7] 0 for normal or MR2[7] 1 for extended
temperature range.
5. READ burst type: Nibble sequential, set MR0[3] 0.
6. In the dual-rank DDP case, note the following IDD measurement considerations:
• For all IDD measurements except IDD6, the unselected rank should be in an IDD2P
condition.
• For all IPP measurements except IPP6, the unselected rank should be in an IDD3N con-
dition.
• For all IDD6/IPP6 measurements, both ranks should be in the same IDD6 condition.
Current Specifications – Patterns and Test Conditions

Current Test Definitions and Patterns
Table 136: IDD0 and IPP0 Measurement-Loop Pattern1
A[17,13,11]]
RAS_n/A16
CAS_n/A15
CK_t, CK_c
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 ACT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 –
1, 2 D, D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
3, 4 D_n, 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
D_n
... Repeat pattern 1...4 until nRAS - 1; truncate if necessary
nRAS PRE 0 1 0 1 0 0 0 0 0 0 0 0 0 0 –
... Repeat pattern 1...4 until nRC - 1; truncate if necessary
1 1 × nRC Repeat sub-loop 0, use BG[1:0] = 1, use BA[1:0] = 1 instead
Static High
Toggling

8 8 × nRC Repeat sub-loop 0, use BG[1:0] = 2, use BA[1:0] = 0 instead4
Notes: 1. DQS_t, DQS_c are VDDQ.

2. BG1 is a "Don't Care" for x16 devices.
3. DQ signals are VDDQ.
4. For x4 and x8 only.
Table 137: IDD1 Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 ACT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 –
1, 2 D, D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
3, 4 D_n, D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
... Repeat pattern 1...4 until nRCD - AL - 1; truncate if necessary
nRCD - AL RD 0 1 1 0 1 0 0 0 0 0 0 0 0 0 D0 = 00, D1 =
... Repeat pattern 1...4 until nRAS - 1; truncate if necessary FF,
D2 = FF, D3 =
nRAS PRE 0 1 0 1 0 0 0 0 0 0 0 0 0 0
00,
... Repeat pattern 1...4 until nRC - 1; truncate if necessary D4 = FF, D5 =
00,
D5 = 00, D7 = FF
1 1 × nRC + 0 ACT 0 0 0 1 1 0 1 1 0 0 0 0 0 0 –
1 × nRC + 1, D, D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
2
1 × nRC + 3, D_n, D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
4
... Repeat pattern nRC + 1...4 until 1 × nRC + nRAS - 1; truncate if necessary
1 × nRC RD 0 1 1 0 1 0 1 1 0 0 0 0 0 0 D0 = FF, D1 =
Static High
Toggling
+nRCD - AL 00,
... Repeat pattern 1...4 until nRAS - 1; truncate if necessary D2 = 00, D3 =
FF,
1 × nRC + PRE 0 1 0 1 0 0 1 1 0 0 0 0 0 0
D4 = 00, D5 =
nRAS
FF,
... Repeat pattern nRC + 1...4 until 2 × nRC - 1; truncate if necessary D5 = FF, D7 = 00
Notes: 1. DQS_t, DQS_c are VDDQ when not toggling.


3. DQ signals are VDDQ except when burst sequence drives each DQ signal by a READ com-
mand.
Table 138: IDD2N, IDD3N, and IPP3P Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
1 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
2 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
3 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
1 4–7 Repeat sub-loop 0, use BG[1:0] = 1, use BA[1:0] = 1 instead
Static High

Toggling

8 32–35 Repeat sub-loop 0, use BG[1:0] = 2, use BA[1:0] = 0 instead4

Table 139: IDD2NT Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
1 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
2 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
3 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
1 4–7 Repeat sub-loop 0 with ODT = 1, use BG[1:0] = 1, use BA[1:0] = 1 instead
Static High
Toggling
8 32–35 Repeat sub-loop 0 with ODT = 0, use BG[1:0] = 2, use BA[1:0] = 0 instead4
Notes: 1. DQS_t, DQS_c are VSSQ.

3. DQ signals are VSSQ.
Table 140: IDD4R Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 RD 0 1 1 0 1 0 0 0 0 0 0 0 0 0 D0 = 00, D1 =
1 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 FF,
D2 = FF, D3 =
2, 3 D_n, 1 1 1 1 1 0 3 3 0 0 0 7 F 0
00,
D_n
D4 = FF, D5 =
00,
D5 = 00, D7 =
FF
1 4 RD 0 1 1 0 1 0 1 1 0 0 0 7 F 0 D0 = FF, D1 = 00
5 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 D2 = 00, D3 =
FF
6, 7 D_n, 1 1 1 1 1 0 3 3 0 0 0 7 F 0
D4 = 00, D5 =
D_n
FF
D5 = FF, D7 = 00
Static High
Toggling


3. Burst sequence driven on each DQ signal by a READ command. Outside burst operation,
DQ signals are VDDQ.
Table 141: IDD4W Measurement – Loop Pattern1
WE_n/A14
RAS_n/A1
CAS_n/A1
Sub-Loop
A[17,13,1
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
mand
Cycle
CK_c,
Com-
CK_t,
CS_n
ODT
CKE
1]]
Data3
5
0 0 WR 0 1 1 0 0 1 0 0 0 0 0 0 0 0 D0 = 00, D1 = FF,
1 D 1 0 0 0 0 1 0 0 0 0 0 0 0 0 D2 = FF, D3 = 00,
D4 = FF, D5 = 00,
2, 3 D_n, 1 1 1 1 0 1 3 3 0 0 0 7 F 0
D5 = 00, D7 = FF
D_n
1 4 WR 0 1 1 0 0 1 1 1 0 0 0 7 F 0 D0 = FF, D1 = 00
5 D 1 0 0 0 0 1 0 0 0 0 0 0 0 0 D2 = 00, D3 = FF
D4 = 00, D5 = FF
6, 7 D_n, 1 1 1 1 0 1 3 3 0 0 0 7 F 0
D5 = FF, D7 = 00
D_n
Static High
Toggling


3. Burst sequence driven on each DQ signal by WRITE command. Outside burst operation,
Table 142: IDD4Wc Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]3
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data4
0 0 WR 0 1 1 0 0 1 0 0 0 0 0 0 0 0 D0 = 00, D1 = FF,
1, 2 D, D 1 0 0 0 0 1 0 0 0 0 0 0 0 0 D2 = FF, D3 = 00,
D4 = FF, D5 = 00,
3, 4 D_n, 1 1 1 1 0 1 3 3 0 0 0 7 F 0
D8 = CRC
D_n
1 5 WR 0 1 1 0 0 1 1 1 0 0 0 7 F 0 D0 = FF, D1 = 00,
6, 7 D, D 1 0 0 0 0 1 0 0 0 0 0 0 0 0 D2 = 00, D3 = FF,
D4 = 00, D5 = FF,
8, 9 D_n, 1 1 1 1 0 1 3 3 0 0 0 7 F 0
D5 = FF, D7 = 00
D_n
D8 = CRC
Static High
Toggling

Notes: 1. Pattern provided for reference only.

2. DQS_t, DQS_c are VDDQ when not toggling.
4. Burst sequence driven on each DQ signal by WRITE command. Outside burst operation,
Table 143: IDD5R Measurement – Loop Pattern1
A[17,13,11]]
CK_c, CK_t,
RAS_n/A16
CAS_n/A15
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 REF 0 1 0 0 1 0 0 0 0 0 0 0 0 0 –
1 1 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
2 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
3 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
4 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
5–8 Repeat pattern 1...4, use BG[1:0] = 1, use BA[1:0] = 1 instead
Static High
Toggling

33–36 Repeat pattern 1...4, use BG[1:0] = 2, use BA[1:0] = 0 instead4
2 65...nREFI - Repeat sub-loop 1; truncate if necessary
1

Table 144: IDD7 Measurement – Loop Pattern1
A[17,13,11]]
RAS_n/A16
CAS_n/A15
CK_t, CK_c
Command
WE_n/A14
Sub-Loop
A12/BC_n
A[10]/AP
BG[1:0]2
Number
BA[1:0]
ACT_n
A[9:7]
A[6:3]
A[2:0]
Cycle
CS_n
ODT
CKE
Data3
0 0 ACT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 –
1 RDA 0 1 1 0 1 0 0 0 0 0 1 0 0 0
2 D 1 0 0 0 0 0 0 0 0 0 0 0 0 0 –
3 D_n 1 1 1 1 1 0 3 3 0 0 0 7 F 0 –
... Repeat pattern 2...3 until nRRD - 1, if nRRD > 4. Truncate if necessary
1 nRRD ACT 0 0 0 0 0 0 1 1 0 0 0 0 0 0 –
nRRD+1 RDA 0 1 1 0 1 0 1 1 0 0 1 0 0 0
... Repeat pattern 2...3 until 2 × nRRD - 1, if nRRD > 4. Truncate if necessary
2 2 × nRRD Repeat sub-loop 0, use BG[1:0] = 0, use BA[1:0] = 2 instead
3 3 × nRRD Repeat sub-loop 1, use BG[1:0] = 1, use BA[1:0] = 3 instead
4 4 × nRRD Repeat pattern 2...3 until nFAW - 1, if nFAW > 4 × nRRD. Truncate if necessary
5 nFAW Repeat sub-loop 0, use BG[1:0] = 0, use BA[1:0] = 1 instead
6 nFAW + nRRD Repeat sub-loop 1, use BG[1:0] = 1, use BA[1:0] = 2 instead
7 nFAW + 2 × nRRD Repeat sub-loop 0, use BG[1:0] = 0, use BA[1:0] = 3 instead
8 nFAW + 3 × nRRD Repeat sub-loop 1, use BG[1:0] = 1, use BA[1:0] = 0 instead
Static High
Toggling
9 nFAW + 4 × nRRD Repeat sub-loop 4

10 2 × nFAW Repeat sub-loop 0, use BG[1:0] = 2, use BA[1:0] = 0 instead
11 2 × nFAW + nRRD Repeat sub-loop 1, use BG[1:0] = 3, use BA[1:0] = 1 instead
12 2 × nFAW + 2 × Repeat sub-loop 0, use BG[1:0] = 2, use BA[1:0] = 2 instead
nRRD
nRRD
14 2 × nFAW + 4 × Repeat sub-loop 4
nRRD
15 3 × nFAW Repeat sub-loop 0, use BG[1:0] = 2, use BA[1:0] = 1 instead
16 3 × nFAW + nRRD Repeat sub-loop 1, use BG[1:0] = 3, use BA[1:0] = 2 instead
nRRD
nRRD
19 3 × nFAW + 4 × Repeat sub-loop 4
nRRD
20 4 × nFAW Repeat pattern 2...3 until nRC - 1, if nRC > 4 × nFAW. Truncate if necessary

3. DQ signals are VDDQ except when burst sequence drives each DQ signal by a READ com-
mand.
IDD Specifications
Table 145: Timings used for IDD, IPP, and IDDQ Measurement – Loop Patterns
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400 DDR4-2666 DDR4-2933 DDR4-3200

10-10-10
11-11-11
12-12-12
12-12-12
13-13-13
14-14-14
14-14-14
15-15-15
16-16-16
16-16-16
17-17-17
18-18-18
18-18-18
19-19-19
20-20-20
20-20-20
21-21-21
22-22-22
20-20-20
22-22-22
24-24-24
Uni
Symbol t
tCK 1.25 1.071 0.937 0.833 0.75 0.682 0.625 ns
CL 10 11 12 12 13 14 14 15 16 16 17 18 18 19 20 20 21 22 20 22 24 CK
CWL 9 11 11 10 12 12 11 14 14 16 16 16 18 18 18 14 18 18 16 20 20 CK
nRCD 10 11 12 12 13 14 14 15 16 16 17 18 18 19 20 19 20 21 20 22 24 CK
nRC 38 39 40 44 45 46 50 51 52 55 56 57 61 62 63 66 67 68 72 74 76 CK
nRP 10 11 12 12 13 14 14 15 16 16 17 18 18 19 20 19 20 21 20 22 24 CK
nRAS 28 32 36 39 43 47 52 CK
nFA x41 16 16 16 16 16 16 16 CK
W x8 20 22 23 26 28 31 34 CK
x1 28 28 32 36 40 44 48 CK
6
nRRD x4 4 4 4 4 4 4 4 CK
_S x8 4 4 4 4 4 4 4 CK
x1 5 6 6 7 8 8 9 CK
6
nRRD x4 5 5 6 6 7 8 8 CK
_L x8 5 5 6 6 7 8 8 CK
x1 6 6 7 8 9 10 11 CK
6
nCCD_S 4 4 4 4 4 4 4 CK
nCCD_L 5 5 6 6 7 8 8 CK
nWTR_S 2 3 3 3 4 4 4 CK
nWTR_L 6 7 8 9 10 11 12 CK
nREFI 6,240 7,283 8,325 9,364 10,400 11,437 12,480 CK
nRFC 2Gb 128 150 171 193 214 235 256 CK
nRFC 4Gb 208 243 278 313 347 382 416 CK
nRFC 8Gb 280 327 374 421 467 514 560 CK
nRFC 280 327 374 421 467 514 560 CK
16Gb
Note: 1. 1KB based x4 use same numbers of clocks for nFAW as the x8.
Current Specifications – Limits
Table 146: IDD, IPP, and IDDQ Current Limits – Rev. B (0°C ื TC ื 95°C)
Symbol Width DDR4-2400 DDR4-2666 Unit

IDD0: One bank ACTIVATE-to-PRECHARGE cur- x8 56 59 mA
rent x16 69 72 mA
IPP0: One bank ACTIVATE-to-PRECHARGE IPP ALL 4 4 mA
current
IDD1: One bank ACTIVATE-to-READ-to- PRE- x8 73 76 mA
CHARGE current x16 97 100 mA
IDD2N: Precharge standby current ALL 41 42 mA
IDD2NT: Precharge standby ODT current x8 50 54 mA
x16 56 58 mA
IDD2P: Precharge power-down current ALL 22 22 mA
IDD2Q: Precharge quiet standby current ALL 36 36 mA
IDD3N: Active standby current ALL 57 58 mA
IPP3N: Active standby IPP current ALL 3 3 mA
IDD3P: Active power-down current ALL 33 33 mA
IDD4R: Burst read current x8 157 169 mA
x16 236 255 mA
IDD4W: Burst write current x8 130 140 mA
x16 179 193 mA
IDD5R: Distributed refresh current (1X REF) ALL 65 66 mA
IPP5R: Distributed refresh IPP current (1X REF) ALL 5 5 mA
IDD6N: Self refresh current1 ALL 24 24 mA
IDD6E: Self refresh current2 ALL 47 47 mA
IDD6R: Self refresh current3, 4 ALL 25 25 mA
IDD6A: Auto self refresh current, 25°C 4 ALL 9 9 mA
IPP6x: Auto self refresh current23 ALL 3 3 mA
IDD7: Bank interleave read current x8 211 225 mA
x16 289 297 mA
IPP7: Bank interleave read IPP current x8 14 1 mA
x16 21 21 mA
IDD8: Maximum power-down current ALL 19 19 mA
Notes: 1. Applicable for MR2 settings A7 = 0 and A6 = 0; manual mode with normal temperature
range of operation (-40–85°C).
2. Applicable for MR2 settings A7 = 1 and A6 = 0; manual mode with extended tempera-
ture range of operation (-40–95°C).
3. Applicable for MR2 settings A7 = 0 and A6 = 1; manual mode with reduced temperature
4. IDD6E, DD6Rand IDD6A values are verified by design and characterization, and may not be
subject to production test.
5. When additive latency is enabled for IDD0, current changes by approximately 9%.
6. When additive latency is enabled for IDD1, current changes by approximately +14% (x8),
+14% (x16).
7. When additive latency is enabled for IDD2N, current changes by approximately 0%.
8. When DLL is disabled for IDD2N, current changes by approximately 1%.
9. When CAL is enabled for IDD2N, current changes by approximately –34%.
10. When gear-down is enabled for IDD2N, current changes by approximately 0%.
11. When CA parity is enabled for IDD2N, current changes by approximately +15%.
12. When additive latency is enabled for IDD3N, current changes by approximately +9%.
13. When additive latency is enabled for IDD4R, current changes by approximately +6%.
14. When read DBI is enabled for IDD4R, current changes by approximately -8%.
15. When additive latency is enabled for IDD4W, current changes by approximately +6% (x8),
+4% (x16).
16. When write DBI is enabled for IDD4W, current changes by approximately 13%.
17. When write CRC is enabled for IDD4W, current changes by approximately +4%.
18. When CA parity is enabled for IDD4W, current changes by approximately +15% (x8),
+10% (x16).
19. When 2X REF is enabled for IDD5R, current changes by approximately –16%.
21. IPP0 test and limit is applicable for IDD0 and IDD1 conditions.
22. IPP3N test and limit is applicable for all IDD2x, IDD3x, IDD4x and IDD8 conditions; that is, test-
ing IPP3N should satisfy the IPPs for the noted IDD tests.
23. IPP6x is applicable to IDD6N, IDD6E, IDD6R and IDD6A conditions.

rent x16 71 74 mA
current
x16 58 60 mA
x16 241 260 mA
x16 184 198 mA
Table 147: IDD, IPP, and IDDQ Current Limits – Rev. B (0°C ื TC ื 105°C) (Continued)

IDD6A: Auto self refresh current ALL 67 67 mA
x16 294 302 mA
x16 23 23 mA
+14% (x16).
+4% (x16).
+10% (x16).

rent x16 76 79 mA
current
x16 63 65 mA
x16 251 270 mA
x16 189 203 mA
IDD6A: Auto self refresh current ALL 87 87 mA
x16 304 312 mA
x16 26 26 mA
+14% (x16).
+4% (x16).
+10% (x16).
Table 149: IDD, IPP, and IDDQ Current Limits – Rev. F(0°C ื TC ื 95°C)
Symbol Width DDR4-2400 DDR4-2666 DDR4-3200 Unit

IDD0: One bank ACTIVATE-to-PRE- x8 43 45 49 mA
CHARGE current x16 50 52 56 mA
IPP0: One bank ACTIVATE-to-PRE- x8 3 3 3 mA
CHARGE IPP current x16 4 4 4
IDD1: One bank ACTIVATE-to-READ-to- x8 59 61 65 mA
PRECHARGE current x16 77 79 83 mA
IDD2N: Precharge standby current ALL 31 32 34 mA
IDD2NT: Precharge standby ODT cur- x8 40 42 46 mA
rent x16 48 51 57 mA
IDD2P: Precharge power-down current ALL 25 25 25 mA
IDD2Q: Precharge quiet standby cur- ALL 27 27 27 mA
rent
IDD3N: Active standby current x8 39 41 45 mA
x16 40 42 46
IPP3N: Active standby IPP current ALL 3 3 3 mA
IDD3P: Active power-down current x8 31 32 34 mA
x16 32 33 35
IDD4R: Burst read current x8 150 161 184 mA
x16 261 282 322 mA
Table 149: IDD, IPP, and IDDQ Current Limits – Rev. F(0°C ื TC ื 95°C) (Continued)

IDD4W: Burst write current x8 127 136 155 mA
x16 205 223 258 mA
IDD5R: Distributed refresh current (1X ALL 54 55 57 mA
REF)
IPP5R: Distributed refresh IPP current ALL 5 5 5 mA
(1X REF)
IDD6N: Self refresh current1 ALL 24 24 24 mA
IDD6E: Self refresh current2 ALL 44 44 44 mA
IDD6R: Self refresh current3, 4 ALL 16 16 16 mA
IDD6A: Auto self refresh current, 25°C 4 ALL 8.6 8.6 8.6 mA
IDD6A: Auto self refresh current, 45°C 4 ALL 16 16 16 mA
IPP6x: Auto self refresh current23 ALL 5 5 5 mA
IDD7: Bank interleave read current x8 175 180 190 mA
x16 243 252 270 mA
IPP7: Bank interleave read IPP current x8 13 13 13 mA
x16 18 18 18 mA
IDD8: Maximum power-down current ALL 18 18 18 mA
+7% (x16).
8. When DLL is disabled for IDD2N, current changes by approximately -6%.
+4% (x16).
16. When write DBI is enabled for IDD4W, current changes by approximately -20%.
17. When write CRC is enabled for IDD4W, current changes by approximately -5%(x8), -5%
(x16).
+12% (x16).
19. When 2X REF is enabled for IDD5R, current changes by approximately 0%.
21. When 2X REF is enabled for IPP5R, current changes by approximately 0%.
26. When Tc < 0°C: IDD2P and IDD3P must be derated by 6%; IDD4R and IDD4W must be derated
by 4%; IDD6, IDD6E and IDD7 must be derated by 11%.

rent x16 49 53 59 mA
rent
x16 42 44 48
x16 35 36 38
x16 265 292 326 mA
x16 205 223 258 mA
REF)
(1X REF)

x16 248 257 275 mA
x16 18 18 18 mA
+7% (x16).
+4% (x16).
(x16).
+12% (x16).

rent x16 50 55 61 mA
rent
x16 44 46 50
x16 38 39 41
x16 269 302 330 mA
x16 215 233 268 mA
REF)
(1X REF)
IDD6A: Auto self refresh current, 95°C 4 ALL mA
x16 253 262 280 mA
x16 18 18 18 mA

+7% (x16).
+4% (x16).
(x16).
+12% (x16).
Speed Bin Tables
Speed Bin Tables

DDR4 DRAM timing is primarily covered by two types of tables: the Speed Bin tables in
this section and the tables found in the Electrical Characteristics and AC Timing Param-
eters section. The timing parameter tables define the applicable timing specifications
based on the speed rating. The Speed Bin tables on the following pages list the tAA,
tRCD, tRP, tRAS, and tRC limits of a given speed mark and are applicable to the CL set-
tings in the lower half of the table provided they are applied in the correct clock range,
which is noted.
Backward Compatibility
Although the speed bin tables list the slower data rates, tAA, CL, and CWL, it is difficult
to determine whether a faster speed bin supports all of the tAA, CL, and CWL combina-
tions across all the data rates of a slower speed bin. To assist in this process, please refer
to the Backward Compatibility table.
CCMTD-1725822587-10418
Table 152: Backward Compatibility

Note 1 applies to the entire table.
Component
Speed Bin Speed Bin Supported
-125 -125E -107 -107E -093 -093E -083D -083 -083E -075D -075 -075E -068D -068 -068E -062 -062E -062Y
-125 yes
-125E yes2 yes
-107 yes yes
-107E yes2 yes yes2 yes
-093 yes yes yes
-093E yes2 yes yes2 yes yes2 yes
-083D yes yes yes yes
-083 yes yes yes yes yes
-083E yes2 yes yes2 yes yes2 yes yes2 yes2 yes
-075D yes yes yes yes yes
-075 yes yes yes yes yes yes yes
339
-075E yes yes yes yes yes yes yes yes yes yes yes

-068D yes yes yes yes yes yes
-068 yes yes yes yes yes yes yes yes yes
-068E yes yes yes yes yes yes yes yes yes yes
-062 yes yes yes yes yes yes yes
-062E yes2 yes2 yes2 yes2 yes2 yes2 yes2 yes2 yes2 yes2 yes
-062Y yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes
Speed Bin Tables

Speed Bin Tables
Notes: 1. The backward compatibility table is not meant to guarantee that any new device will be
a drop in replacement for an existing part number. Customers should review the operat-
ing conditions for any device to determine its suitability for use in their design.
2. This condition exceeds the JEDEC requirement in order to allow additional flexibility for
components. However, JEDEC SPD compliance may force modules to only support the JE-
DEC-defined value. Refer to the SPD documentation for further clarification.
CCMTD-1725822587-10418
Table 153: DDR4-1600 Speed Bins and Operating Conditions

Notes 1–3 apply to the entire table
DDR4-1600 Speed Bin -125E -125
CL-nRCD-nRP 11-11-11 12-12-12
Parameter Symbol Min Max Min Max Unit
Internal READ command to first data tAA 13.75 19.006 15.00 19.006 ns
(13.50)4
Internal READ command to first data with read DBI enabled tAA_DBI tAA tAA tAA tAA ns
(MIN) + (MAX) + (MIN) + (MAX) +
2nCK 2nCK 2nCK 2nCK
ACTIVATE-to-internal READ or WRITE delay time tRCD 13.75 – 15.00 – ns
(13.50)4
PRECHARGE command period tRP 13.75 – 15.00 – ns
(13.50)4
ACTIVATE-to-PRECHARGE command period tRAS 35 9 × tREFI 35 9 × tREFI ns
ACTIVATE-to-ACTIVATE or REFRESH command period tRC5 tRAS + – tRAS + – ns
tRP tRP
Data Rate Equivalent tAAmin(ns): READ CL: READ WRITE

341
Symbol Min Max Min Max Unit

Max (MT/s) Speed Bin non-DB nonDBI CL: DBI CWL
1333 - 13.50 9 11 9 tCK (AVG) 1.500 1.9006 Reserved ns

- 15.00 10 12 tCK (AVG) 1.5006 1.9006 1.500 1.9006 ns
1600 -125E 13.75 11 13 9, 11 tCK (AVG) 1.250 <1.500 Reserved ns

-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
Supported CL settings 9, 106, 11-12 10, 12 nCK
Supported CL settings with read DBI 11, 126, 13-14 12, 14 nCK
Supported CWL settings 9, 11 9, 11 nCK
Speed Bin Tables

Speed Bin Tables
Notes: 1. Speed Bin table is only valid with DLL enabled.

2. When operating in 2tCK WRITE preamble mode, CWL must be programmed to a value
range.
3. The programmed value of CWL must be less than or equal to the programmed value of
CL.
4. This value applies to non-native tCK-CL-nRCD-nRP combinations.
5. When calculating tRC in clocks, values may not be used in a combination that violate
tRAS or tRP.
6. This value exceeds the JEDEC requirement in order to allow additional flexibility, espe-
cially for components. However, JEDEC SPD compliance may force modules to only sup-
port the JEDEC defined value, please refer to the SPD documentation.
CCMTD-1725822587-10418

CL-nRCD-nRP 13-13-13 14-14-14
(13.50)4
Internal READ command to first data with read DBI enabled tAA_DBI tAA (MIN) tAA tAA (MIN) tAA ns
+ 2nCK (MAX) + + 2nCK (MAX) +
2nCK 2nCK
ACTIVATE to internal READ or WRITE delay time tRCD 13.92 – 15.00 – ns
(13.50)4
(13.50)4
tRP tRP
Data Rate Equivalent tAAmin: non- READ CL: READ CL: WRITE
343

Max (MT/s) Speed Bin DBI nonDBI DBI CWL
1333 – 13.50 9 11 9 tCK (AVG) 1.500 1.9006 Reserved ns

– 15.00 10 12 tCK (AVG) 1.5006 1.9006 1.500 1.9006 ns

-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
-107 15.00 14 16 tCK (AVG) 1.071 <1.250 ns
Supported CL settings 9, 106 , 11–14 10, 12, 14 nCK
Supported CL settings with read DBI 11, 126 , 13–16 12, 14, 16 nCK
Supported CWL settings 9–12 9–12 nCK
Speed Bin Tables

Speed Bin Tables

range.
CL.
tRAS or tRP.
CCMTD-1725822587-10418

CL-nRCD-nRP 15-15-15 16-16-16
(13.50)4
Internal READ command to first data with read DBI enabled tAA_DBI tAA (MIN) tAA tAA (MIN) tAA ns
+ 3nCK (MAX) + + 3nCK (MAX) +
3nCK 3nCK
ACTIVATE to internal READ or WRITE delay time tRCD 14.06 – 15.00 – ns
(13.50)4
(13.50)4
tRP tRP
Data Rate Equivalent tAAmin (ns): READ CL: READ CL: WRITE
345

Max (MT/s) Speed Bin non-DBI non-DBI DBI CWL
1333 – 13.50 9 11 9 tCK (AVG) 1.500 1.9006 Reserved ns

– 15.00 10 12 tCK (AVG) 1.5006 1.9006 1.500 1.9006 ns
1600 -125E 13.75 11 13 9 , 11 tCK (AVG) 1.250 <1.500 Reserved ns

-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
-107 15.00 14 16 tCK (AVG) 1.071 <1.250 ns
-093 15.00 16 19 tCK (AVG) 0.937 <1.071 ns
Supported CL settings 9, 106 , 11–16 10, 12, 14, 16 nCK
Supported CL settings with read DBI 11, 126 , 13–16, 18-19 12, 14, 16, 19 nCK
Speed Bin Tables

Supported CWL settings 9, 10, 11, 12, 14 9, 10, 11, 12, 14 nCK
Speed Bin Tables

range.
CL.
tRAS or tRP.
CCMTD-1725822587-10418

DDR4-2400 Speed Bin -083E -083 -083D
CL-nRCD-nRP 16-16-16 17-17-17 18-18-18
Internal READ command to first data tAA 13.32 19.006 14.16 19.006 15.00 19.006 ns
(13.75)4
Internal READ command to first data with read DBI enabled tAA_DBI tAA tAA tAA tAA tAA tAA ns
(MIN) + (MAX) + (MIN) + (MAX) + (MIN) + (MAX) +
3nCK 3nCK 3nCK 3nCK 3nCK 3nCK
ACTIVATE to internal READ or WRITE delay time tRCD 13.32 – 14.16 – 15.00 19.00 ns
(13.75)4
PRECHARGE command period tRP 13.32 – 14.16 – 15.00 19.00 ns
(13.75)4
ACTIVATE-to-PRECHARGE command period tRAS 32 9 × tREFI 32 9 × tREFI 32 32 ns
ACTIVATE-to-ACTIVATE or REFRESH command period tRC5 tRAS + – tRAS + – tRAS + – ns
tRP tRP tRP
tAAmin READ READ

347
Data Rate Equivalent WRITE

(ns): CL: CL: Symbol Min Max Min Max Min Max Unit
Max (MT/s) Speed Bin CWL
non-DBI non-DBI DBI

1333 – 13.50 9 11 9 tCK (AVG) 1.500 1.9006 Reserved Reserved ns
– 15.00 10 12 tCK (AVG) 1.5006 1.9006 1.500 1.9006 1.500 1.9006 ns

1600 -125E 13.75 11 13 9, 11 tCK (AVG) 1.250 <1.500 1.250 <1.500 Reserved ns
-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
-107 15.00 14 16 tCK (AVG) 1.071 <1.250 ns
-093 15.00 16 19 tCK (AVG) 0.937 <1.071 ns
2400 -083E 13.32 16 19 12, 16 tCK (AVG) 0.833 <0.937 Reserved Reserved ns
Speed Bin Tables

-083 14.16 17 20 tCK (AVG) 0.833 <0.937 ns
-083D 15.00 18 21 tCK (AVG) 0.833 <0.937 ns
Supported CL settings 9, 106 , 11–18 10–18 10, 12, 14, 16, 18 nCK
Supported CL settings with read DBI 11, 126 , 13–16, 12–16, 18–21 12, 14, 16, 19, 21 nCK
18–21
Supported CWL settings 9–12, 14, 16 9-12, 14, 16 9–12, 14, 16 nCK
Speed Bin Tables

range.
CL.
tRAS or tRP.
CCMTD-1725822587-10418

DDR4-2666 Speed Bin -075E -075 -075D
CL-nRCD-nRP 18-18-18 19-19-19 20-20-20
Internal READ command to first data tAA 14.25
13.50 19.006 19.006 15.00 19.006 ns
(13.75)4
Internal READ command to first data with read DBI enabled tAA tAA tAA tAA tAA tAA
tAA_DBI (MIN) + (MAX) + (MIN) + (MAX) + (MIN) + (MAX) + ns
ACTIVATE to internal READ or WRITE delay time tRCD 14.25
13.50 – – 15.00 – ns
(13.75)4
PRECHARGE command period tRP 14.25
13.50 – – 15.00 – ns
(13.75)4
ACTIVATE-to-PRECHARGE command period tRAS 32 9 × tREFI 32 9 × tREFI 32 9 × tREFI ns
ACTIVATE-to-ACTIVATE or REFRESH command period tRAS + tRAS + tRAS +
tRC5 – – – ns
tRP tRP tRP
tAAmin READ READ

349

non-DBI non-DBI DBI

1333 - 13.50 9 11 tCK (AVG) Reserved Reserved ns
9 1.500 1.9006
- 15.00 10 12 tCK (AVG) 1.500 1.9006 1.500 1.9006 ns

1600 -125E 13.75 11 13 tCK (AVG) Reserved ns
9, 11 tCK
1.250 <1.500 1.250 <1.500
-125 15.00 12 14 (AVG) 1.250 <1.500 ns
10, 12 tCK
1.071 <1.250 1.071 <1.250
-107 15.00 14 16 (AVG) 1.071 <1.250 ns
11, 14 tCK
0.937 <1.071 0.937 <1.071
-093 15.00 16 19 (AVG) 0.937 <1.071 ns
2400 -083E 13.32 16 19 tCK (AVG) Reserved Reserved ns

Reserved
Speed Bin Tables

-083 14.16 17 20 12, 16 tCK (AVG) ns
tCK
0.833 <0.937 0.833 <0.937
-083D 15.00 18 21 (AVG) 0.833 <0.937 ns
tCK
Reserved
-075 14.25 19 22 14, 18 (AVG) 0.750 <0.833 ns
tCK
0.750 <0.833
-075D 15.00 20 23 (AVG) 0.750 <0.833 ns
Supported CL settings 9–20 10-20 10, 12, 14, 16, 18, nCK
20
CCMTD-1725822587-10418
Table 157: DDR4-2666 Speed Bins and Operating Conditions (Continued)

DDR4-2666 Speed Bin -075E -075 -075D
CL-nRCD-nRP 18-18-18 19-19-19 20-20-20
Supported CL settings with read DBI 11–16, 18–23 12–16, 18–23 12, 14, 16, 19, 21, nCK
23
Supported CWL settings 9–12, 14, 16, 18 9–12, 14, 16, 18 9–12, 14, 16, 18 nCK
350

Speed Bin Tables

Speed Bin Tables

range.
CL.
tRAS or tRP.
CCMTD-1725822587-10418

DDR4-2933 Speed Bin -068E -068 -068D
CL-nRCD-nRP 20-20-20 21-21-21 22-22-22
(13.75)4
ACTIVATE-to-internal READ or WRITE delay time tRCD 13.64 – 14.32 – 15.00 – ns
(13.75)4
PRECHARGE command period tRP 13.64 – 14.32 – 15.00 – ns
(13.75)4
tRP tRP tRP
READ READ
352
Data Rate Equivalent tAAmin(ns): WRITE

CL: CL: Symbol Min Max Min Max Min Max Unit
Max (MT/s) Speed Bin non-DBI CWL
non-DBI DBI

1333 – 13.50 9 11 9 tCK (AVG) Reserved Reserved Reserved ns
– 15.00 10 12 tCK (AVG) 1.500 1.9006 1.500 1.9006 1.500 1.9006 ns

-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
-107 15.00 14 16 tCK (AVG) 1.071 <1.250 ns
-093 15.00 16 19 tCK (AVG) 0.937 <1.071 ns
2400 -083E 13.32 16 19 12, 16 tCK (AVG) Reserved Reserved Reserved ns
Speed Bin Tables

-083 14.16 17 20 tCK (AVG) 0.833 <0.937 0.833 <0.937 ns
083D 15.00 18 21 tCK (AVG) 0.833 <0.937 ns
-075 14.25 19 22 tCK (AVG) 0.750 <0.833 0.750 <0.833 ns
-075D 15.00 20 23 tCK (AVG) 0.750 <0.833 ns
CCMTD-1725822587-10418

DDR4-2933 Speed Bin -068E -068 -068D
CL-nRCD-nRP 20-20-20 21-21-21 22-22-22
-068 14.32 21 25 tCK (AVG) 0.682 <0.750 ns
-068D 15.00 22 26 tCK (AVG) 0.682 <0.750 ns
– 16.37 24 28 tCK (AVG) Reserved Reserved Reserved ns
Supported CL settings 10–22 10–22 10, 12, 14, 16, 18, nCK
20, 22
Supported CL settings with read DBI 12–16, 18–26 12–16,18–23, 12, 14, 16, 19, 21, nCK
25-26 23, 26
Supported CWL settings 9–12, 14, 16, 18, 9–12, 14, 16, 18, 9–12, 14, 16, 18, nCK
20 20 20
353

Speed Bin Tables

Speed Bin Tables

range.
CL.
tRAS or tRP.
CCMTD-1725822587-10418

DDR4-3200 Speed Bin -062Y6 -062E -062
CL-nRCD-nRP 22-22-22 22-22-22 24-24-24
(13.32)4
ACTIVATE-to-internal READ or WRITE delay time tRCD 13.75 – 13.75 – 15.00 – ns
(13.32)4
PRECHARGE command period tRP 13.75 – 13.75 – 15.00 – ns
(13.32)4
tRP tRP tRP
tAAmin READ READ

355

non-DBI non-DBI DBI

1333 - 13.50 9 11 9 tCK (AVG) 1.500 1.9006 Reserved Reserved ns
- 15.00 10 12 tCK (AVG) 1.500 1.9006 1.500 1.9006 ns

-125 15.00 12 14 tCK (AVG) 1.250 <1.500 ns
-107 15.00 14 16 tCK (AVG) 1.071 <1.250 ns
-093 15.00 16 19 tCK (AVG) 0.937 <1.071 ns
Speed Bin Tables

-083 14.16 17 20 tCK (AVG) 0.833 <0.937 ns
-083D 15.00 18 21 tCK (AVG) 0.833 <0.937 ns
-075 14.25 19 22 tCK (AVG) 0.750 <0.833 ns
-075D 15.00 20 23 tCK (AVG) 0.750 <0.833 ns
CCMTD-1725822587-10418

DDR4-3200 Speed Bin -062Y6 -062E -062
CL-nRCD-nRP 22-22-22 22-22-22 24-24-24
-068 14.32 21 25 tCK (AVG) 0.682 <0.750 0.682 <0.750 ns
-068D 15.00 22 26 tCK (AVG) 0.682 <0.750 0.682 <0.750 ns
– 16.37 24 28 tCK (AVG) 0.682 <0.750 ns
-062 15.00 24 28 tCK (AVG) 0.625 <0.682 ns
Supported CL settings 9–22, 24 10–22, 24 10, 12, 14, 16, 18, nCK
20, 22, 24
Supported CL settings with read DBI 11–16, 18–23, 12–16, 18–23, 12, 14, 16, 19, 21, nCK
25-26, 28 25-26, 28 23, 26, 28
Supported CWL settings 9–12, 14, 16, 18, 9–12, 14, 16, 18, 9–12, 14, 16, 18, nCK
20 20 20
356

Speed Bin Tables

Refresh Parameters By Device Density

range.
CL.
tRAS or tRP.
Refresh Parameters By Device Density
Table 160: Refresh Parameters by Device Density
Parameter Symbol 2Gb 4Gb 8Gb 16Gb Unit Notes

REF command to ACT or REF tRFC (All bank groups) 160 260 350 550 ns
command time
Average periodic refresh inter- tREFI -40°C ื TC ื 85°C 7.8 7.8 7.8 3.9 μs
val 85°C < TC ื 95°C 3.9 3.9 3.9 1.95 μs 1
95°C < TC ื 105°C 1.95 1.95 1.95 0.975 μs
105°C < TC ื 0.975 0.975 0.975 0.4875 μs 1
125°C
Note: 1. Users should refer to the DRAM supplier data sheet and/or the DIMM SPD to determine
if the devices support these options or requirements.
CCMTD-1725822587-10418
Electrical Characteristics and AC Timing Parameters – DDR4-1600 through DDR4-2400
Table 161: Electrical Characteristics and AC Timing Parameters
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

Clock Timing
Clock period average (DLL off mode) tCK (DLL_OFF) 8 20 8 20 8 20 8 20 ns
Clock period average tCK (AVG, 1.25 1.9 1.071 1.9 0.937 1.9 0.833 1.9 ns 14
DLL_ON)
High pulse width average tCH (AVG) 0.48 0.52 0.48 0.52 0.48 0.52 0.48 0.52 CK
Low pulse width average tCL (AVG) 0.48 0.52 0.48 0.52 0.48 0.52 0.48 0.52 CK
Clock period jitter Total tJITper_tot –63 63 –54 54 –47 47 –42 42 ps
Deterministic tJITper_tot –31 31 –27 27 –23 23 –21 21 ps
Electrical Characteristics and AC Timing Parameters –

DLL locking tJITper,lck –50 50 –43 43 –38 38 -33 33 ps
Clock absolute period tCK (ABS) MIN = tCK (AVG) MIN + tJITper_tot MIN; MAX = tCK (AVG) MAX + ps
tJITper_tot MAX
Clock absolute high pulse width tCH (ABS) 0.45 – 0.45 – 0.45 – 0.45 – tCK
358
(includes duty cycle jitter) (AVG)

Clock absolute low pulse width tCL (ABS) 0.45 – 0.45 – 0.45 – 0.45 – tCK

Cycle-to-cycle jitter Total tJITcc _tot - 125 - 107 - 94 - 83 ps
DLL locking tJITcc,lck - 100 - 86 - 75 - 67 ps

Cumulative error across 2 cycles tERR2per –92 92 –79 79 –69 69 –61 61 ps
DDR4-1600 through DDR4-2400

3 cycles tERR3per –109 109 –94 94 –82 82 –73 73 ps
n = 13, 14 . . . 49, tERRnper tERRnper MIN = (1 + 0.68ln[n]) × tJITper_tot MIN ps
50 cycles tERRnper MAX = (1 + 0.68ln[n]) × tJITper_tot MAX
CCMTD-1725822587-10418
Table 161: Electrical Characteristics and AC Timing Parameters (Continued)
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

DQ Input Timing
Data setup time to Base (calibrated tDS Refer to DQ Input Receiver Specification section –
DQS_t, DQS_c VREF) (approximately 0.15tCK to 0.28tCK )
Noncalibrated tPDA_S minimum of 0.5UI UI
VREF
Data hold time from Base (calibrated tDH Refer to DQ Input Receiver Specification section –
Noncalibrated tPDA_H minimum of 0.5UI UI
VREF
DQ and DM minimum data pulse width tDIPW 0.58 – 0.58 – 0.58 – 0.58 – UI

for each input
DQ Output Timing (DLL enabled)
DQS_t, DQS_c to DQ skew, per group, per tDQSQ – 0.16 – 0.16 – 0.16 – 0.17 UI
access
tQH
359
DQ output hold time from DQS_t, DQS_c 0.76 – 0.76 – 0.76 – 0.74 – UI
Data Valid Window per device: tQH - tDVW
d 0.63 0.63 0.64 0.64 UI

tDQSQ each device’s output per UI
Data Valid Window per device, per pin: tDVW 0.66 - 0.66 - 0.69 - 0.72 - UI
p
tQH - tDQSQ each device’s output per UI

DQ Low-Z time from CK_t, CK_c tLZDQ –450 225 –390 195 –360 180 –330 175 ps

DQ High-Z time from CK_t, CK_c tHZDQ – 225 – 195 – 180 – 175 ps
DQ Strobe Input Timing
DQS_t, DQS_c rising edge to CK_t, CK_c tDQSS –0.27 0.27 –0.27 0.27 –0.27 0.27 –0.27 0.27 CK
1ck
rising edge for 1tCK preamble
DQS_t, DQS_c rising edge to CK_t, CK_c tDQSS –0.50 0.50 –0.50 0.50 –0.50 0.50 –0.50 0.50 CK
2ck
rising edge for 2tCK preamble
DQS_t, DQS_c differential input low pulse tDQSL 0.46 0.54 0.46 0.54 0.46 0.54 0.46 0.54 CK
width
DQS_t, DQS_c differential input high tDQSH 0.46 0.54 0.46 0.54 0.46 0.54 0.46 0.54 CK
pulse width
DQS_t, DQS_c falling edge setup to CK_t, tDSS 0.18 – 0.18 – 0.18 – 0.18 – CK
CK_c rising edge
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

DQS_t, DQS_c falling edge hold from tDSH 0.18 – 0.18 – 0.18 – 0.18 – CK
CK_t, CK_c rising edge
DQS_t, DQS_c differential WRITE pream- tWPRE 0.9 – 0.9 – 0.9 – 0.9 – CK
1ck
ble for 1tCK preamble
DQS_t, DQS_c differential WRITE pream- tWPRE 1.8 – 1.8 – 1.8 – 1.8 – CK
2ck
DQS_t, DQS_c differential WRITE postam- tWPST 0.33 – 0.33 – 0.33 – 0.33 – CK
ble
DQS Strobe Output Timing (DLL enabled)
DQS_t, DQS_c rising edge output access tDQSCK –225 225 –195 195 –180 180 –175 175 ps

time from rising CK_t, CK_c
DQS_t, DQS_c rising edge output var- tDQSCKi - 370 - 330 - 310 - 290 ps
iance window per DRAM
DQS_t, DQS_c differential output high tQSH 0.4 – 0.4 – 0.4 – 0.4 – CK
time
360
DQS_t, DQS_c differential output low tQSL 0.4 – 0.4 – 0.4 – 0.4 – CK
time

DQS_t, DQS_c Low-Z time (RL - 1) tLZDQS –450 225 –390 195 –360 180 –330 175 ps
DQS_t, DQS_c High-Z time (RL + BL/2) tHZDQS – 225 – 195 – 180 – 175 ps
DQS_t, DQS_c differential READ pream- tRPRE 0.9 – 0.9 – 0.9 – 0.9 – CK
1ck

DQS_t, DQS_c differential READ pream- tRPRE 1.8 – 1.8 – 1.8 – 1.8 – CK
2ck
DQS_t, DQS_c differential READ postam- tRPST 0.33 – 0.33 – 0.33 – 0.33 – CK
ble
Command and Address Timing
DLL locking time tDLLK 597 – 597 – 768 – 768 – CK 2, 4

CMD, ADDR setup time Base tIS 115 – 100 – 80 – 62 – ps
to CK_t, CK_c Base ref- VREFCA tIS
VREF 215 – 200 – 180 – 162 – ps
erenced to VIH(AC) and
VIL(AC) levels
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

CMD, ADDR hold time Base tIH 140 – 125 – 105 – 87 – ps
to CK_t, CK_c Base ref- VREFCA tIH
VREF 215 – 200 – 180 – 162 – ps
erenced to VIH(DC) and
VIL(DC) levels
CTRL, ADDR pulse width for each input tIPW 600 – 525 – 460 – 410 – ps
ACTIVATE to internal READ or WRITE de- tRCD See Speed Bin Tables for tRCD ns
lay
PRECHARGE command period tRP See Speed Bin Tables for tRP ns
ACTIVATE-to-PRECHARGE command peri- tRAS See Speed Bin Tables for tRAS ns 13
od

ACTIVATE-to-ACTIVATE or REF command tRC See Speed Bin Tables for tRC ns 13
period
ACTIVATE-to-ACTIVATE command period tRRD_S MIN = greater MIN = greater MIN = greater MIN = greater CK 1
to different bank groups for 1/2KB page (1/2KB) of 4CK or 5ns of 4CK or 4.2ns of 4CK or 3.7ns of 4CK or 3.3ns
size
361
to different bank groups for 1KB page (1KB) of 4CK or 5ns of 4CK or 4.2ns of 4CK or 3.7ns of 4CK or 3.3ns

size
to different bank groups for 2KB page (2KB) of 4CK or 6ns of 4CK or 5.3ns of 4CK or 5.3ns of 4CK or 5.3ns
size

ACTIVATE-to-ACTIVATE command period tRRD_L MIN = greater MIN = greater MIN = greater MIN = greater CK 1
to same bank groups for 1/2KB page size (1/2KB) of 4CK or 6ns of 4CK or 5.3ns of 4CK or 5.3ns of 4CK or 4.9ns
to same bank groups for 1KB page size (1KB) of 4CK or 6ns of 4CK or 5.3ns of 4CK or 5.3ns of 4CK or 4.9ns
to same bank groups for 2KB page size (2KB) of 4CK or 7.5ns of 4CK or 6.4ns of 4CK or 6.4ns of 4CK or 6.4ns
Four ACTIVATE windows for 1/2KB page tFAW MIN = greater MIN = greater MIN = greater MIN = greater ns
size (1/2KB) of 16CK or 20ns of 16CK or 17ns of 16CK or 15ns of 16CK or 13ns
Four ACTIVATE windows for 1KB page tFAW MIN = greater MIN = greater MIN = greater MIN = greater ns
size (1KB) of 20CK or 25ns of 20CK or 23ns of 20CK or 21ns of 20CK or 21ns
Four ACTIVATE windows for 2KB page tFAW MIN = greater MIN = greater MIN = greater MIN = greater ns
size (2KB) of 28CK or 35ns of 28CK or 30ns of 28CK or 30ns of 28CK or 30ns
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

WRITE recovery time tWR MIN = 15ns ns 6, 10, 1
tWR MIN = 1CK + tWR CK 6, 11, 1
2
WRITE recovery time when CRC and DM tWR_CRC_DM MIN = tWR_L + MIN = tWR_L + greater of (5CK or 3.75ns) CK 7, 10, 1
are both enabled greater of (4CK
or 3.75ns)
tWR_CRC_DM MIN = 1CK + tWR_CRC_DM CK 7, 11, 1
2
Delay from start of internal WRITE trans- tWTR_L MIN = greater of 4CK or 7.5ns CK 6, 10, 1
action to internal READ command – Same tWTR_L
2 MIN = 1CK + tWTR_L CK 6, 11, 1
bank group
Delay from start of internal WRITE trans- tWTR_L_CRC_D MIN = tWTR_L + MIN = tWTR_L + greater of (5CK or 3.75ns) CK 7, 10, 1

action to internal READ command – Same M greater of (4CK
bank group when CRC and DM are both or 3.75ns)
enabled tWTR_L_CRC_D MIN = 1CK + tWTR_L_CRC_DM CK 7, 11, 1
M2
Delay from start of internal WRITE trans- tWTR_S MIN = greater of (2CK or 2.5ns) CK 6, 8, 9,
362
action to internal READ command – Dif- 10, 1

ferent bank group tWTR_S MIN = 1CK + tWTR_S CK 6, 8, 9,

2
11, 1
Delay from start of internal WRITE trans- tWTR_S_CRC_D MIN = tWTR_S + MIN = tWTR_S + greater of (5CK or 3.75ns) CK 7, 8, 9,
action to internal READ command – Dif- M greater of (4CK 10, 1

ferent bank group when CRC and DM are or 3.75ns)
both enabled tWTR_S_CRC_D MIN = 1CK + tWTR_S_CRC_DM CK 7, 8, 9,
M2 11, 1
READ-to-PRECHARGE time tRTP MIN = greater of 4CK or 7.5ns CK 1
CAS_n-to-CAS_n command delay to dif- tCCD_S 4 – 4 – 4 – 4 – CK
ferent bank group
CAS_n-to-CAS_n command delay to same tCCD_L MIN = – MIN = – MIN = – MIN = – CK 2, 15

bank group greater greater greater greater
of 4CK of 4CK of 4CK of 4CK
or or or or 5ns
6.25ns 5.355ns 5.355ns
Auto precharge write recovery + pre- tDAL (MIN) MIN = WR + ROUNDUPtRP/tCK (AVG); MAX = N/A CK
charge time
MRS Command Timing
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

MRS command cycle time tMRD 8 – 8 – 8 – 8 – CK
MRS command cycle time in PDA mode tMRD_PDA MIN = greater of (16nCK, 10ns) CK 1
MRS command cycle time in CAL mode tMRD_CAL MIN = tMOD + tCAL CK
MRS command update delay in PDA tMOD MIN = greater of (24nCK, 15ns) CK 1
mode
MRS command update delay tMOD_PDA MIN = tMOD CK
MRS command update delay in CAL tMOD_CAL MIN = tMOD + tCAL CK
mode
MRS command to DQS drive in preamble tSDO MIN = tMOD + 9ns
training

MPR Command Timing
Multipurpose register recovery time tMPR MIN = 1CK CK
Multipurpose register write recovery time tWR_MPR MIN = tMOD + AL + PL
CRC Error Reporting Timing
363
CRC error to ALERT_n latency tCRC_ALERT 3 13 3 13 3 13 3 13 ns

CRC ALERT_n pulse width tCRC_ALERT_P 6 10 6 10 6 10 6 10 CK

W
CA Parity Timing
Parity latency PL 4 – 4 – 4 – 5 – CK

Commands uncertain to be executed dur- tPAR_UN- - PL - PL - PL - PL CK
ing this time KNOWN
Delay from errant command to ALERT_n tPAR_ALERT_O – PL + – PL+ – PL + – PL + CK
assertion N 6ns 6ns 6ns 6ns
Pulse width of ALERT_n signal when as- tPAR_ALERT_P 48 96 56 112 64 128 72 144 CK
serted W
Time from alert asserted until DES com- tPAR_ALERT_RS – 43 – 50 – 57 – 64 CK

mands required in persistent CA parity P
mode
CAL Timing
CS_n to command address latency tCAL 3 – 4 – 4 – 5 – CK
CS_n to command address latency in tCALg N/A – N/A – N/A – N/A – CK
gear-down mode
MPSM Timing
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

Command path disable delay upopn tMPED MIN = tMOD (MIN) + tCPDED (MIN) CK 1
MPSM entry
Valid clock requirement after MPSM tCKMPE MIN = tMOD (MIN) + tCPDED (MIN) CK 1
entry
Valid clock requirement before MPSM tCKMPX MIN = tCK - tCKSRX (MIN) CK 1
exit
Exit MPSM to commands not requiring a tXMP tXS (MIN) CK
locked DLL
Exit MPSM to commands requiring a tXMPDLL MIN = tXMP (MIN) + tXSDLL (MIN) CK 1
locked DLL

CS setup time to CKE tMPX_S MIN = tIS (MIN) + tIH (MIN) ns
CS_n HIGH hold time to CKE rising edge tMPX_HH MIN = tXP ns
CS_n LOW hold time to CKE rising edge tMPX_LH 12 tXMP-1 12 tXMP-1 12 tXMP-1 12 tXMP-1 ns
0ns 0ns 0ns 0ns
364
Connectivity Test Timing

TEN pin HIGH to CS_n LOW – Enter CT tCT_Enable 200 – 200 – 200 – 200 – ns

mode
CS_n LOW and valid input to valid output tCT_Valid – 200 – 200 – 200 – 200 ns
CK_t, CK_c valid and CKE HIGH after TEN tCTCKE_Valid 10 – 10 – 10 – 10 – ns

goes HIGH

Calibration and VREFDQ Train Timing
ZQCL command: Long POWER-UP and tZQinit 1024 – 1024 – 1024 – 1024 – CK
calibration time RESET operation
Normal opera- tZQoper 512 – 512 – 512 – 512 – CK
tion
tZQCS
ZQCS command: Short calibration time 128 – 128 – 128 – 128 – CK

The VREF increment/decrement step time VREF_time MIN = 150ns
Enter VREFDQ training mode to the first tVREFDQE MIN = 150ns ns 1
write or VREFDQ MRS command delay
Exit VREFDQ training mode to the first tVREFDQX MIN = 150ns ns 1
WRITE command delay
Initialization and Reset Timing
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

Exit reset from CKE HIGH to a valid com- tXPR MIN = greater of 5CK or tRFC (MIN) + 10ns CK 1
mand
RESET_L pulse low after power stable tPW_REST_S 0.1 – 0.1 – 0.1 – 0.1 – μs
RESET_L pulse low at power-up tPW_REST_L 200 – 200 – 200 – 200 – μs
Begin power supply ramp to power sup- tVDDPR MIN = N/A; MAX = 200 ms
plies stable
RESET_n LOW to power supplies stable tRPS MIN = 0; MAX = 0 ns
RESET_n LOW to I/O and RTT High-Z tIOZ MIN = N/A; MAX = undefined ns
Refresh Timing

REFRESH-to-ACTIVATE tRFC1 MIN = 260 ns 1, 12
or REFRESH command 4Gb tRFC2 MIN = 160 ns 1, 12
period (all bank tRFC4 MIN = 110 ns 1, 12
groups)
tRFC1 MIN = 350 ns 1, 12
8Gb tRFC2 MIN = 260 ns 1, 12
365
tRFC4 MIN = 160 ns 1, 12

tRFC1

MIN = 550 ns 1, 12
16Gb tRFC2 MIN = 350 ns 1, 12
tRFC4 MIN = 260 ns 1, 12

Average periodic re- -40°C ื TC ื 85°C tREFI MIN = N/A; MAX = 7.8 μs 12

fresh interval 85°C < TC ื 95°C tREFI MIN = N/A; MAX = 3.9 μs 12
95°C < TC ื tREFI MIN = N/A; MAX = 1.95 μs 12
105°C
125°C
Self Refresh Timing

Exit self refresh to commands not requir-
ing a locked DLL tXS MIN = tRFC + 10ns ns 1
Exit self refresh to commands not requir- tXS_ABORT MIN = tRFC4 + 10ns
ing a locked DLL in self refresh abort ns 1
Exit self refresh to ZQCL, ZQCS and MRS tXS_FAST MIN = tRFC4 + 10ns 1
(CL, CWL, WR, RTP and gear-down) ns
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

Exit self refresh to commands requiring a tXSDLL MIN = tDLLK (MIN) CK 1
locked DLL
Minimum CKE low pulse width for self re- tCKESR MIN = tCKE (MIN) + 1nCK CK 1
fresh entry to self refresh exit timing
Minimum CKE low pulse width for self re- tCKESR_PAR MIN = tCKE (MIN) + 1nCK + PL CK 1
when CA parity is enabled
Valid clocks after self refresh entry (SRE) tCKSRE MIN = greater of (5CK, 10ns) CK 1
or power-down entry (PDE)
Valid clock requirement after self refresh tCKSRE_PAR MIN = greater of (5CK, 10ns) + PL CK 1

entry or power-down when CA parity is
enabled
Valid clocks before self refresh exit (SRX) tCKSRX MIN = greater of (5CK, 10ns) CK 1
or power-down exit (PDX), or reset exit
Power-Down Timing
366
Exit power-down with DLL on to any val- tXP MIN = greater of 4CK or 6ns CK 1
id command

Exit power-down with DLL on to any val- tXP _PAR MIN = (greater of 4CK or 6ns) + PL CK 1
id command when CA Parity is enabled.

CKE MIN pulse width tCKE (MIN) MIN = greater of 3CK or 5ns CK 1

Command pass disable delay tCPDED 4 – 4 – 4 – 4 – CK
Power-down entry to power-down exit tPD MIN = tCKE (MIN); MAX = 9 × tREFI CK
timing
Begin power-down period prior to CKE tANPD WL - 1CK CK
registered HIGH
Power-down entry period: ODT either PDE Greater of tANPD or tRFC - REFRESH command to CKE LOW time CK
synchronous or asynchronous
Power-down exit period: ODT either syn- PDX tANPD + tXSDLL CK
chronous or asynchronous
Power-Down Entry Minimum Timing
ACTIVATE command to power-down en- tACTPDEN 1 – 1 – 2 – 2 – CK
try
PRECHARGE/PRECHARGE ALL command tPRPDEN 1 – 1 – 2 – 2 – CK
to power-down entry
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

REFRESH command to power-down entry tREFPDEN 1 – 1 – 2 – 2 – CK
MRS command to power-down entry tMRSPDEN MIN = tMOD (MIN) CK 1
READ/READ with auto precharge com- tRDPDEN MIN = RL + 4 + 1 CK 1
mand to power-down entry
WRITE command to power-down entry tWRPDEN MIN = WL + 4 + tWR/tCK(AVG) CK 1
(BL8OTF, BL8MRS, BC4OTF)
WRITE command to power-down entry tWRPBC4DEN MIN = WL + 2 + tWR/tCK(AVG) CK 1
(BC4MRS)
WRITE with auto precharge command to tWRAPDEN MIN = WL + 4 + WR + 1 CK 1
power-down entry (BL8OTF,

BL8MRS,BC4OTF)
WRITE with auto precharge command to tWRAPBC4DEN MIN = WL + 2 + WR + 1 CK 1
power-down entry (BC4MRS)
ODT Timing
367
Direct ODT turn-on latency DODTLon WL - 2 = CWL + AL + PL - 2 CK

Direct ODT turn-off latency DODTLoff WL - 2 = CWL + AL + PL - 2 CK

RTT dynamic change skew tADC 0.3 0.7 0.3 0.7 0.3 0.7 0.3 0.7 CK
Asynchronous RTT(NOM) turn-on delay tAONAS 1 9 1 9 1 9 1 9 ns
(DLL off)
Asynchronous RTT(NOM) turn-off delay tAOFAS 1 9 1 9 1 9 1 9 ns

(DLL off)
ODT HIGH time with WRITE command ODTH8 1tCK 6 – 6 – 6 – 6 – CK
and BL8 ODTH8 2tCK 7 – 7 – 7 – 7 –
ODT HIGH time without WRITE command ODTH4 1tCK 4 – 4 – 4 – 4 – CK
or with WRITE command and BC4 ODTH4 2tCK 5 – 5 – 5 – 5 –
Write Leveling Timing

First DQS_t, DQS_c rising edge after write tWLMRD 40 – 40 – 40 – 40 – CK
leveling mode is programmed
DQS_t, DQS_c delay after write leveling tWLDQSEN 25 – 25 – 25 – 25 – CK
mode is programmed
Write leveling setup from rising CK_t, tWLS 0.13 – 0.13 – 0.13 – 0.13 – CK(AV
CK_c crossing to rising DQS_t, DQS_c G)
crossing
CCMTD-1725822587-10418
DDR4-1600 DDR4-1866 DDR4-2133 DDR4-2400

Write leveling hold from rising DQS_t, tWLH 0.13 – 0.13 – 0.13 – 0.13 – CK(AV
DQS_c crossing to rising CK_t, CK_c cross- G)
ing
Write leveling output delay tWLO 0 9.5 0 9.5 0 9.5 0 9.5 ns
Write leveling output error tWLOE 0 2 0 2 0 2 0 2 ns
Gear-Down Timing (Not Supported Below DDR4-2666)
Exit reset from CKE HIGH to a valid MRS tXPR_GEAR N/A N/A N/A N/A CK
gear-down
CKE HIGH assert to gear-down enable tXS_GEAR N/A N/A N/A N/A CK
time)

MRS command to sync pulse time tSYNC_GEAR N/A N/A N/A N/A CK
Sync pulse to first valid command tCMD_GEAR N/A N/A N/A N/A CK
Gear-down setup time tGEAR_setup N/A – N/A – N/A – N/A – CK
Gear-down hold time tGEAR_hold N/A – N/A – N/A – N/A – CK
368


CCMTD-1725822587-10418
Notes: 1. Maximum limit not applicable.
2. tCCD_L and tDLLK should be programmed according to the value defined per operating frequency.
3. Although unlimited row accesses to the same row is allowed within the refresh period, excessive row accesses to
the same row over a long term can result in degraded operation.
4. Data rate is greater than or equal to 1066 Mb/s.
5. RFUcorr
6. WRITE-to-READ when CRC and DM are both not enabled.
7. WRITE-to-READ delay when CRC and DM are both enabled.
8. The start of internal write transactions is defined as follows:
• For BL8 (fixed by MRS and on-the-fly): rising clock edge four clock cycles after WL
• For BC4 (on-the-fly): rising clock edge four clock cycles after WL
• For BC4 (fixed by MRS): rising clock edge two clock cycles after WL
9. For these parameters, the device supports tnPARAM [nCK] = RU{tPARAM [ns]/tCK(AVG) [ns]}, in clock cycles, assum-
ing all input clock jitter specifications are satisfied.
10. When operating in 1tCK WRITE preamble mode.

12. When CA parity mode is selected and the DLLoff mode is used, each REF command requires an additional "PL"
added to tRFC refresh time.
13. DRAM devices should be evenly addressed when being accessed. Disproportionate accesses to a particular row ad-
dress may result in reduction of the product lifetime and/or reduction in data retention ability.
369
14. Applicable from tCK (AVG) MIN to tCK (AVG) MAX as stated in the Speed Bin tables.
15. JEDEC specifies a minimum of five clocks.

16. The maximum read postamble is bound by tDQSCK (MIN) plus tQSH (MIN) on the left side and tHZ(DQS) MAX on
the right side.
17. The reference level of DQ output signal is specified with a midpoint as a widest part of output signal eye, which
should be approximately 0.7 × VDDQ as a center level of the static single-ended output peak-to-peak swing with

a driver impedance of 34 ohms and an effective test load of 50 ohms to VTT = VDDQ.
18. JEDEC hasn't agreed upon the definition of the deterministic jitter; the user should focus on meeting the total
limit.
19. Spread spectrum is not included in the jitter specification values. However, the input clock can accommodate
spread-spectrum at a sweep rate in the range of 20–60 kHz with an additional 1% of tCK (AVG) as a long-term
jitter component; however, the spread spectrum may not use a clock rate below tCK (AVG) MIN.
20. The actual tCAL minimum is the larger of 3 clocks or 3.748ns/tCK; the table lists the applicable clocks required at
targeted speed bin.

21. The maximum READ preamble is bounded by tLZ(DQS) MIN on the left side and tDQSCK (MAX) on the right side.
See figure in the Clock to Data Strobe Relationship section. Boundary of DQS Low-Z occur one cycle earlier in 2tCK
toggle mode, which is illustrated in the READ Preamble section.
22. DQ falling signal middle-point of transferring from HIGH to LOW to first rising edge of DQS differential signal
cross-point.
CCMTD-1725822587-10418
Electrical Characteristics and AC Timing Parameters – DDR4-2666 through DDR4-3200
Table 162: Electrical Characteristics and AC Timing Parameters
DDR4-2666 DDR4-2933 DDR4-3200 Reserved

Clock Timing
Clock period average (DLL off mode) tCK (DLL_OFF) 8 20 8 20 8 20 ns
Clock period average tCK (AVG, 0.75 1.9 0.682 1.9 0.625 1.9 ns 14
DLL_ON)
High pulse width average tCH (AVG) 0.48 0.52 0.48 0.52 0.48 0.52 CK
Low pulse width average tCL (AVG) 0.48 0.52 0.48 0.52 0.48 0.52 CK
Clock period jitter Total tJITper_tot –38 38 -34 34 –32 32 ps
Deterministic tJITper_dj –19 19 -17 17 –16 16 ps

DLL locking tJITper,lck –30 30 -27 27 –25 25 ps
Clock absolute period tCK (ABS) MIN = tCK (AVG) MIN + tJITper_tot MIN; MAX = tCK (AVG) MAX + ps
tJITper_tot MAX
Clock absolute high pulse width tCH (ABS) 0.45 – 0.45 – 0.45 – tCK
370

Clock absolute low pulse width tCL (ABS) 0.45 – 0.45 – 0.45 – tCK

Cycle-to-cycle jitter Total tJITcc _tot 75 68 62 ps
DLL locking tJITcc,lck 60 55 50 ps

Cumulative error across 2 cycles tERR2per –55 55 -50 50 –46 46 ps

3 cycles tERR3per –66 66 -60 60 –55 55 ps

n = 13, 14 . . . 49, tERRnper tERRnper MIN = (1 + 0.68ln[n]) × tJITper_tot MIN ps
50 cycles tERRnper MAX = (1 + 0.68ln[n]) × tJITper_tot MAX
CCMTD-1725822587-10418

DQ Input Timing
Data setup time to Base (calibrated tDS Refer to DQ Input Receiver Specification section –
Non-calibrated tPDA_S minimum of 0.5ui UI
VREF
Data hold time from Base (calibrated tDH Refer to DQ Input Receiver Specification section –
Non-calibrated tPDA_H minimum of 0.5UI UI
VREF
DQ and DM minimum data pulse width tDIPW 0.58 – 0.58 – 0.58 – UI

for each input
DQ Output Timing (DLL enabled)
DQS_t, DQS_c to DQ skew, per group, per tDQSQ – 0.18 – 0.19 – 0.22 UI
access
tQH
371
DQ output hold time from DQS_t, DQS_c 0.74 – 0.74 – 0.74 – UI

Data Valid Window per device: tQH - tDVW
d 0.64 – 0.64 – 0.64 – UI

tDQSQ each device’s output per IU
Data Valid Window per device, per pin: tDVW 0.72 – 0.72 – 0.72 – UI
p
tQH - tDQSQ each device’s output per IU

DQ Low-Z time from CK_t, CK_c tLZDQ –310 170 –280 165 –250 160 ps

DQ High-Z time from CK_t, CK_c tHZDQ – 170 – 165 – 160 ps
DQ Strobe Input Timing
DQS_t, DQS_c rising edge to CK_t, CK_c tDQSS –0.27 0.27 –0.27 0.27 –0.27 0.27 CK
1ck
rising edge for 1tCKpreamble
DQS_t, DQS_c rising edge to CK_t, CK_c tDQSS –0.50 0.50 –0.50 0.50 –0.50 0.50 CK
2ck
rising edge for 2tCKpreamble
DQS_t, DQS_c differential input low pulse tDQSL 0.46 0.54 0.46 0.54 0.46 0.54 CK
width
DQS_t, DQS_c differential input high tDQSH 0.46 0.54 0.46 0.54 0.46 0.54 CK
pulse width
DQS_t, DQS_c falling edge setup to CK_t, tDSS 0.18 – 0.18 – 0.18 – CK
CK_c rising edge
CCMTD-1725822587-10418

DQS_t, DQS_c falling edge hold from tDSH 0.18 – 0.18 – 0.18 – CK
CK_t, CK_c rising edge
DQS_t, DQS_c differential WRITE pream- tWPRE 0.9 – 0.9 – 0.9 – CK
1ck
ble for 1tCKpreamble
DQS_t, DQS_c differential WRITE pream- tWPRE 1.8 – 1.8 – 1.8 – CK
2ck
DQS_t, DQS_c differential WRITE postam- tWPST 0.33 – 0.33 – 0.33 – CK
ble
DQS Strobe Output Timing (DLL enabled)
DQS_t, DQS_c rising edge output access tDQSCK –170 170 –165 165 –160 160 ps

time from rising CK_t, CK_c
DQS_t, DQS_c rising edge output var- tDQSCKi – 270 – 265 – 260 ps
iance window per DRAM
DQS_t, DQS_c differential output high tQSH 0.40 – 0.40 – 0.40 – CK
time
372
DQS_t, DQS_c differential output low tQSL 0.40 – 0.40 – 0.40 – CK

time

DQS_t, DQS_c Low-Z time (RL - 1) tLZDQS –310 170 –280 165 –250 160 ps
DQS_t, DQS_c High-Z time (RL + BL/2) tHZDQS – 170 – 165 – 160 ps
DQS_t, DQS_c differential READ pream- tRPRE 0.9 – 0.9 – 0.9 – CK
1ck

DQS_t, DQS_c differential READ pream- tRPRE 1.8 – 1.8 – 1.8 – CK
2ck
DQS_t, DQS_c differential READ postam- tRPST 0.33 – 0.33 – 0.33 – CK
ble
Command and Address Timing
DLL locking time tDLLK 854 – 940 – 1024 – CK 2, 3

CMD, ADDR setup time Base tIS 55 – 48 – 40 – ps
to CK_t, CK_c refer- VREFCA tIS
VREF 145 – 138 – 130 – ps
enced to VIH(AC) and
VIL(AC) levels
CCMTD-1725822587-10418

CMD, ADDR hold time Base tIH 80 – 73 – 65 – ps
to CK_t, CK_c refer- VREFCA tIH
VREF 145 – 138 – 130 – ps
enced to VIH(DC) and
VIL(DC) levels
CTRL, ADDR pulse width for each input tIPW 385 – 365 – 350 – ps
ACTIVATE to internal READ or WRITE de- tRCD See Speed Bin Tables for tRCD ns
lay
PRECHARGE command period tRP See Speed Bin Tables for tRP ns
ACTIVATE-to-PRECHARGE command peri- tRAS See Speed Bin Tables for tRAS ns 13
od

ACTIVATE-to-ACTIVATE or REF command tRC See Speed Bin Tables for tRC ns 13
period
ACTIVATE-to-ACTIVATE command period tRRD_S MIN = greater MIN = greater MIN = greater CK 1
to different bank groups for 1/2KB page (1/2KB) of 4CK or 3.0ns of 4CK or 2.7ns of 4CK or 2.5ns
size
373
to different bank groups for 1KB page (1KB) of 4CK or 3.0ns of 4CK or 2.7ns of 4CK or 2.5ns

size
to different bank groups for 2KB page (2KB) of 4CK or 5.3ns of 4CK or 5.3ns of 4CK or 5.3ns
size

ACTIVATE-to-ACTIVATE command period tRRD_L MIN = greater MIN = greater MIN = greater CK 1
to same bank groups for 1/2KB page size (1/2KB) of 4CK or 4.9ns of 4CK or 4.9ns of 4CK or 4.9ns
to same bank groups for 1KB page size (1KB) of 4CK or 4.9ns of 4CK or 4.9ns of 4CK or 4.9ns
to same bank groups for 2KB page size (2KB) of 4CK or 6.4ns of 4CK or 6.4ns of 4CK or 6.4ns
Four ACTIVATE windows for 1/2KB page tFAW MIN = greater MIN = greater MIN = greater ns
size (1/2KB) of 16CK or 12ns of 16CK or of 16CK or 10ns
10.9ns
Four ACTIVATE windows for 1KB page tFAW MIN = greater MIN = greater MIN = greater ns
size (1KB) of 20CK or 21ns of 20CK or 21ns of 20CK or 21ns
Four ACTIVATE windows for 2KB page tFAW MIN = greater MIN = greater MIN = greater ns
size (2KB) of 28CK or 30ns of 28CK or 30ns of 28CK or 30ns
CCMTD-1725822587-10418

WRITE recovery time tWR MIN = 15ns ns 5, 10, 1
tWR MIN = 1CK + tWR CK 5, 11, 1
2
WRITE recovery time when CRC and DM tWR_CRC_DM MIN = tWR + greater of (5CK or 3.75ns) CK 6, 10, 1
are both enabled
WRITE recovery time when CRC and DM tWR_CRC_DM MIN = 1CK + tWR_CRC_DM CK 6, 11, 1
2
are both enabled
Delay from start of internal WRITE trans- tWTR_L MIN = greater of 4CK or 7.5ns CK 5, 10, 1
action to internal READ command – Same tWTR_L
2 MIN = 1CK + tWTR_L CK 5, 11, 1
bank group
Delay from start of internal WRITE trans- tWTR_L_CRC_D MIN = tWTR_L + greater of (5CK or 3.75ns) CK 6, 10, 1

action to internal READ command – Same M
bank group when CRC and DM are both tWTR_L_CRC_D MIN = 1CK + tWTR_L_CRC_DM CK 6, 11, 1
enabled M2
Delay from start of internal WRITE trans- tWTR_S MIN = greater of (2CK or 2.5ns) CK 5, 7, 8,
action to internal READ command – Dif- 10, 1
374
ferent bank group tWTR_S

2 MIN = 1CK + tWTR_S CK 5, 7, 8,
11, 1

Delay from start of internal WRITE trans- tWTR_S_CRC_D MIN = tWTR_S + greater of (5CK or 3.75ns) CK 6, 7, 8,
action to internal READ command – Dif- M 10, 1

ferent bank group when CRC and DM are tWTR_S_CRC_D MIN = 1CK + tWTR_S_CRC_DM CK 6, 7, 8,

both enabled M2 11, 1
READ-to-PRECHARGE time tRTP MIN = greater of 4CK or 7.5ns CK 1
CAS_n-to-CAS_n command delay to dif- tCCD_S 4 – 4 – 4 – CK
ferent bank group
CAS_n-to-CAS_n command delay to same tCCD_L MIN = – MIN = – MIN = – CK 2, 15
bank group greater greater greater
of 4CK of 4CK of 4CK

or 5ns or 5ns or 5ns
Auto precharge write recovery + pre- tDAL (MIN) MIN = WR + ROUNDUPtRP/tCK (AVG); MAX = N/A CK
charge time
MRS Command Timing
MRS command cycle time tMRD 9 – 10 – 10 – CK
MRS command cycle time in PDA mode tMRD_PDA MIN = greater of (16nCK, 10ns) 1
MRS command cycle time in CAL mode tMRD_CAL MIN = tMOD + tCAL CK
CCMTD-1725822587-10418

MRS command update delay in PDA tMOD MIN = greater of (24nCK, 15ns) CK 1
mode
MRS command update delay tMOD_PDA MIN = tMOD CK
MRS command update delay in CAL tMOD_CAL MIN = tMOD + tCAL CK
mode
MRS commandto DQS drive in preamble tSDO MIN = tMOD + 9ns
training
MPR Command Timing
Multipurpose register recovery time tMPR MIN = 1nCK CK
Multipurpose register write recovery time tWR_MPR MIN = tMOD + AL + PL

CRC Error Reporting Timing
CRC error to ALERT_n latency tCRC_ALERT 3 13 3 13 3 13 ns
CRC ALERT_n pulse width tCRC_ALERT_P 6 10 6 10 6 10 CK
W
375
CA Parity Timing
Parity latency PL 5 – 6 – 6 – CK

Commands uncertain to be executed dur- tPAR_UN- – PL – PL – PL CK
ing this time KNOWN
Delay from errant command to ALERT_n tPAR_ALERT_O – PL + – PL + – PL + CK

assertion N 6ns 6ns 6ns

Pulse width of ALERT_n signal when as- tPAR_ALERT_P 80 160 88 176 96 192 CK
serted W
Time from alert asserted until DES com- tPAR_ALERT_RS – 71 – 78 – 85 CK
mands required in persistent CA parity P
mode
CAL Timing
CS_n to command address latency tCAL 5 – 6 – 6 – CK
CS_n to command address latency in tCALg 6 – 8 – 8 – CK
gear-down mode
MPSM Timing
Command path disable delay upopn tMPED MIN = tMOD (MIN) + tCPDED (MIN) CK 1
MPSM entry
CCMTD-1725822587-10418

Valid clock requirement after MPSM tCKMPE MIN = tMOD (MIN) + tCPDED (MIN) CK 1
entry
Valid clock requirement before MPSM tCKMPX MIN = tCK - tCKSRX (MIN) CK 1
exit
Exit MPSM to commands not requiring a tXMP tXS (MIN) CK
locked DLL
Exit MPSM to commands requiring a tXMPDLL MIN = tXMP (MIN) + tXSDLL (MIN) CK 1
locked DLL
CS setup time to CKE tMPX_S MIN = tIS (MIN) + tIH (MIN) ns
CS_n HIGH hold time to CKE rising edge tMPX_HH MIN = tXP ns

CS_n LOW hold time to CKE rising edge tMPX_LH 12 tXMP-1 12 tXMP-1 12 tXMP-1 ns
0ns 0ns 0ns
Connectivity Test Timing
TEN pin HIGH to CS_n LOW – Enter CT tCT_Enable 200 – 200 – 200 – ns
376
mode
CS_n LOW and valid input to valid output tCT_Valid – 200 – 200 – 200 ns

CK_t, CK_c valid and CKE HIGH after TEN tCTCKE_Valid 10 – 10 – 10 – ns
goes HIGH
Calibration and VREFDQ Train Timing

ZQCL command: Long POWER-UP and tZQinit 1024 – 1024 – 1024 – CK

calibration time RESET operation
Normal opera- tZQoper 512 – 512 – 512 – CK
tion
ZQCS command: Short calibration time tZQCS 128 – 128 – 128 – CK
The VREF increment/decrement step time VREF_time MIN = 150ns
Enter VREFDQ training mode to the first tVREFDQE MIN = 150ns ns 1

write or VREFDQ MRS command delay
Exit VREFDQ training mode to the first tVREFDQX MIN = 150ns ns 1
WRITE command delay
Initialization and Reset Timing
Exit reset from CKE HIGH to a valid com- tXPR MIN = greater of 5CK or tRFC (MIN) + 10ns CK 1
mand
RESET_L pulse low after power stable tPW_REST_S 0.1 – 0.1 – 0.1 – μs
CCMTD-1725822587-10418

RESET_L pulse low at power-up tPW_REST_L 200 – 200 – 200 – μs
Begin power supply ramp to power sup- tVDDPR MIN = N/A; MAX = 200 ms
plies stable
RESET_n LOW to power supplies stable tRPS MIN = 0; MAX = 0 ns
RESET_n LOW to I/O and RTT High-Z tIOZ MIN = N/A; MAX = undefined ns
Refresh Timing
REFRESH-to-ACTIVATE tRFC1 MIN = 260 ns 1, 12
or REFRESH command 4Gb tRFC2 MIN = 160 ns 1, 12
period (all bank tRFC4 MIN = 110 ns 1, 12
groups)

tRFC1 MIN = 350 ns 1, 12
8Gb tRFC2 MIN = 260 ns 1, 12
tRFC4 MIN = 160 ns 1, 12
tRFC1 MIN = 550 ns 1, 12
16Gb tRFC2 MIN = 350 ns 1, 12
377
tRFC4 MIN = 260 ns 1, 12

-40°C ื TC ื 85°C

Average periodic re- tREFI MIN = N/A; MAX = 7.8 μs 12
fresh interval 85°C < TC ื 95°C tREFI MIN = N/A; MAX = 3.9 μs 12

105°C

125°C
Self Refresh Timing
Exit self refresh to commands not requir-
ing a locked DLL tXS MIN = tRFC + 10ns ns 1
Exit self refresh to commands not requir- tXS_ABORT MIN = tRFC4 + 10ns
ing a locked DLL in self refresh abort ns 1
Exit self refresh to ZQCL, ZQCS and MRS tXS_FAST MIN = tRFC4 + 10ns 1
(CL, CWL, WR, RTP and gear-down) ns
Exit self refresh to commands requiring a tXSDLL MIN = tDLLK (MIN) CK 1
locked DLL
Minimum CKE low pulse width for self re- tCKESR MIN = tCKE (MIN) + 1nCK CK 1
CCMTD-1725822587-10418

Minimum CKE low pulse width for self re- tCKESR_par MIN = tCKE (MIN) + 1nCK + PL CK 1
when CA parity is enabled
Valid clocks after self refresh entry (SRE) tCKSRE MIN = greater of (5CK, 10ns) CK 1
or power-down entry (PDE)
Valid clock requirement after self refresh tCKSRE_par MIN = greater of (5CK, 10ns) + PL CK 1
entry or power-down when CA parity is
enabled
Valid clocks before self refresh exit (SRX) tCKSRX MIN = greater of (5CK, 10ns) CK 1
or power-down exit (PDX), or reset exit

Power-Down Timing
Exit power-down with DLL on to any val- tXP MIN = greater of 4CK or 6ns CK 1
id command
Exit precharge power-down with DLL fro- tXP _PAR MIN = (greater of 4CK or 6ns) + PL CK 1
zen to commands not requiring a locked
378
DLL when CA Parity is enabled.

CKE MIN pulse width tCKE (MIN) MIN = greater of 3CK or 5ns CK 1

Command pass disable delay tCPDED 4 – 4 – 4 – CK
Power-down entry to power-down exit tPD MIN = tCKE (MIN); MAX = 9 × tREFI CK
timing

Begin power-down period prior to CKE tANPD WL - 1CK CK
registered HIGH
Power-down entry period: ODT either PDE Greater of tANPD or tRFC - REFRESH command to CKE LOW time CK
synchronous or asynchronous
Power-down exit period: ODT either syn- PDX tANPD + tXSDLL CK
chronous or asynchronous
Power-Down Entry Minimum Timing

ACTIVATE command to power-down en- tACTPDEN 2 – 2 – 2 – CK
try
PRECHARGE/PRECHARGE ALL command tPRPDEN 2 – 2 – 2 – CK
to power-down entry
REFRESH command to power-down entry tREFPDEN 2 – 2 – 2 – CK
MRS command to power-down entry tMRSPDEN MIN = tMOD (MIN) CK 1
CCMTD-1725822587-10418

READ/READ with auto precharge com- tRDPDEN MIN = RL + 4 + 1 CK 1
mand to power-down entry
WRITE command to power-down entry tWRPDEN MIN = WL + 4 + tWR/tCK(AVG) CK 1
(BL8OTF, BL8MRS, BC4OTF)
WRITE command to power-down entry tWRPBC4DEN MIN = WL + 2 + tWR/tCK(AVG) CK 1
(BC4MRS)
WRITE with auto precharge command to tWRAPDEN MIN = WL + 4 + WR + 1 CK 1
power-down entry (BL8OTF,
BL8MRS,BC4OTF)
WRITE with auto precharge command to tWRAPBC4DEN MIN = WL + 2 + WR + 1 CK 1

power-down entry (BC4MRS)
ODT Timing
Direct ODT turn-on latency DODTLon WL - 2 = CWL + AL + PL - 2 CK
Direct ODT turn-off latency DODTLoff WL - 2 = CWL + AL + PL - 2 CK
tADC
379
RTT dynamic change skew 0.3 0.7 0.3 0.7 0.3 0.7 CK
Asynchronous RTT(NOM) turn-on delay tAONAS 1 9 1 9 1 9 ns

(DLL off)
Asynchronous RTT(NOM) turn-off delay tAOFAS 1 9 1 9 1 9 ns
(DLL off)
ODT HIGH time with WRITE command ODTH8 1tCK 6 – 6 – 6 – CK

and BL8 ODTH8 2tCK 7 – 7 – 7 –
ODT HIGH time without WRITE command ODTH4 1tCK 4 – 4 – 4 – CK
or with WRITE command and BC4 ODTH4 2tCK 5 – 5 – 5 –
Write Leveling Timing
First DQS_t, DQS_c rising edge after write tWLMRD 40 – 40 – 40 – CK
leveling mode is programmed

DQS_t, DQS_c delay after write leveling tWLDQSEN 25 – 25 – 25 – CK
mode is programmed
Write leveling setup from rising CK_t, tWLS 0.13 – 0.13 – 0.13 – CK
CK_c crossing to rising DQS_t, DQS_c
crossing
CCMTD-1725822587-10418

Write leveling hold from rising DQS_t, tWLH 0.13 – 0.13 – 0.13 – CK
DQS_c crossing to rising CK_t, CK_c cross-
ing
Write leveling output delay tWLO 0 9.5 0 9.5 0 9.5 ns
Write leveling output error tWLOE 0 2 0 2 0 2 ns
Gear-Down Timing
Exit reset from CKE HIGH to a valid MRS tXPR_GEAR tXPR tXPR tXPR CK
gear-down
CKE HIGH assert to gear-down enable tXS_GEAR tXS tXS tXS CK
time)

MRS command to sync pulse time tSYNC_GEAR tMOD + 4CK tMOD + 4CK tMOD + 4CK CK
Sync pulse to first valid command tCMD_GEAR tMOD tMOD tMOD CK
Gear-down setup time tGEAR_setup 2CK – 2CK – 2CK – CK
Gear-down hold time tGEAR_hold 2CK – 2CK – 2CK – CK
380


CCMTD-1725822587-10418
Notes: 1. Maximum limit not applicable.
2. tCCD_L and tDLLK should be programmed according to the value defined per operating frequency.
3. Data rate is greater than or equal to 1066 Mb/s.
4. RFU
5. WRITE-to-READ when CRC and DM are both not enabled.
6. WRITE-to-READ delay when CRC and DM are both enabled.
7. The start of internal write transactions is defined as follows:
• For BL8 (fixed by MRS and on-the-fly): rising clock edge four clock cycles after WL
• For BC4 (on-the-fly): rising clock edge four clock cycles after WL
• For BC4 (fixed by MRS): rising clock edge two clock cycles after WL
8. For these parameters, the device supports tnPARAM [nCK] = RU{tPARAM [ns]/tCK(AVG) [ns]}, in clock cycles, assum-
ing all input clock jitter specifications are satisfied.
9. Although unlimited row accesses to the same row is allowed within the refresh period, excessive row accesses to
the same row over a long term can result in degraded operation.

12. When CA parity mode is selected and the DLLoff mode is used, each REF command requires an additional "PL"
added to tRFC refresh time.
13. DRAM devices should be evenly addressed when being accessed. Disproportionate accesses to a particular row ad-
dress may result in reduction of the product lifetime and/or reduction in data retention ability.
381
14. Applicable from tCK (AVG) MIN to tCK (AVG) MAX as stated in the Speed Bin tables.
15. JEDEC specifies a minimum of five clocks.

16. The maximum read postamble is bound by tDQSCK(MIN) plus tQSH(MIN) on the left side and tHZ(DQS)MAX on the
right side.
17. The reference level of DQ output signal is specified with a midpoint as a widest part of output signal eye, which
should be approximately 0.7 × VDDQ as a center level of the static single-ended output peak-to-peak swing with a

driver impedance of 34 ohms and an effective test load of 50 ohms to VTT = VDDQ.
18. JEDEC hasn't agreed upon the definition of the deterministic jitter; the user should focus on meeting the total
limit.
19. Spread spectrum is not included in the jitter specification values. However, the input clock can accommodate
spread-spectrum at a sweep rate in the range of 20–60 kHz with an additional 1% of tCK (AVG) as a long-term
jitter component; however, clock rate below tCK (AVG) MIN.
20. The actual tCAL minimum is the larger of 3 clocks or 3.748ns/tCK; the table lists the applicable clocks required at
targeted speed bin.

21. The maximum READ preamble is bounded by tLZ(DQS) MIN on the left side and tDQSCK (MAX) on the right side.
See figure in Clock to Data Strobe Relationship. Boundary of DQS Low-Z occur one cycle earlier in 2tCK toggle
mode which is illustrated in READ Preamble.
22. DQ falling signal middle-point of transferring from High to Low to first rising edge of DQS differential signal
cross-point.
Converting Time-Based Specifications to Clock-Based Require-
ments
Converting Time-Based Specifications to Clock-Based Requirements

Software algorithms for calculation of timing parameters are subject to potential round-
ing errors when converting DRAM timing requirements to system clocks; for example, a
memory clock with a nominal frequency of 933.33...3MHz which yields a clock period of
1.071428571429...ns. It is unrealistic to represent all digits after the decimal point exact-
ly and some sort of rounding needs to be done.
DDR4 SDRAM SPD-based specifications use a minimum granularity for SPD-associated
timing parameters of 1ps. Clock periods such as tCK (AVG) MIN are defined to the near-
est picosecond. For example, 1.071428571429...ns is stated as 1071ps. Parameters such
as tAA MIN are specified in units of time (nanoseconds) and require mathematical com-
putation to convert to system clocks (nCK). Rules for rounding allow optimization of
device performance without violating device parameters. These SPD algorithms rely on
results that are within nCK adjustment factors on device testing and specification to
avoid losing performance due to rounding errors when using SPD-based parameters.
Note that JEDEC also defines an nCK adjustment factor, but mandates the inverse nCK
adjustment factor be used in case of conflicting results, so only the inverse nCK adjust-
ment factor is discussed here.
Guidance converting SPD associated timing parameters to system clock requirements:
• Round the application clock period up to the nearest picosecond.
• Express the timing specification and application clock period in picoseconds; scaling
a nanosecond-based parameter value by 1000 allows programmers to use integer
math instead of real math by expressing timing in ps.
• Divide the picosecond-based parameter by the picoseconds based application clock
period.
• Add an inverse nCK adjustment factor of 97.4%.
• Truncate down to the next lower integer value.
• nCK = Truncate[(parameter in ps)/(application tCK in ps) + (974/1000)].
Guidance converting nonSPD associated timing parameters to system clock require-
ments:
• Divide the time base specification (in ns) and divided by the clock period (in ns).
• The resultant is set to the next higher integer number of clocks.
• nCK = Ceiling[(parameter in ns/application tCK in ns)].
Options Tables
Options Tables
Table 163: Options – Speed Based
Data Rate
Function Acronym 1600 1866 2133 2400 2666 2933 3200
Write leveling WL Yes Yes Yes Yes Yes Yes Yes
Temperature controlled refresh TCR Yes Yes Yes Yes Yes Yes Yes
Low-power auto self refresh LPASR Yes Yes Yes Yes Yes Yes Yes
Fine granularity refresh FGR Yes Yes Yes Yes Yes Yes Yes
Multipurpose register MR Yes Yes Yes Yes Yes Yes Yes
Data mask DM Yes Yes Yes Yes Yes Yes Yes
Data bus inversion DBI Yes Yes Yes Yes Yes Yes Yes
TDQS – Yes Yes Yes Yes Yes Yes Yes
ZQ calibration ZQ CAL Yes Yes Yes Yes Yes Yes Yes
VREFDQ calibration – Yes Yes Yes Yes Yes Yes Yes
Per-DRAM addressability Per DRAM Yes Yes Yes Yes Yes Yes Yes
Mode register readout – Yes Yes Yes Yes Yes Yes Yes
Command/Address latency CAL Yes Yes Yes Yes Yes Yes Yes
Write CRC CRC Yes Yes Yes Yes Yes Yes Yes
CA parity – Yes Yes Yes Yes Yes Yes Yes
Gear-down mode – No No No No Yes Yes Yes
Programmable preamble – No No No Yes Yes Yes Yes
Maximum power saving mode MPSM Yes Yes Yes Yes Yes Yes Yes
Additive latency AL Yes Yes Yes Yes Yes Yes Yes
Connectivity test mode CT Yes Yes Yes Yes Yes Yes Yes
Hard post package repair mode hPPR Yes Yes Yes Yes Yes Yes Yes
Soft post package repair mode sPPR Yes Yes Yes Yes Yes Yes Yes
Options Tables
Table 164: Options – Width Based
Width
Function Acronym x4 x8 x16
Write leveling WL Yes Yes Yes
Temperature controlled refresh TCR Yes Yes Yes
Low-power auto self refresh LPASR Yes Yes Yes
Fine granularity refresh FGR Yes Yes Yes
Multipurpose register MR Yes Yes Yes
Data mask DM No Yes Yes
Data bus inversion DBI No Yes Yes
TDQS – No Yes No
ZQ calibration ZQ CAL Yes Yes Yes
VREFDQ calibration – Yes Yes Yes
Per-DRAM addressability Per DRAM Yes Yes Yes
Mode regsiter readout – Yes Yes Yes
Command/Address latency CAL Yes Yes Yes
Write CRC CRC Yes Yes Yes
CA parity – Yes Yes Yes
Gear-down mode – Yes Yes Yes
Programmable preamble – Yes Yes Yes
Maximum power-down mode MPSM Yes Yes No
Additive latency AL Yes Yes Yes
Connectivity test mode CT JEDEC optional on 8Gb and larger densities Yes
Micron supports on all densities
Hard post package repair mode hPPR JEDEC optional on 4Gb
Soft post package repair mode sPPR JEDEC optional on 4Gb and 8Gb
8000 S. Federal Way, P.O. Box 6, Boise, ID 83707-0006, Tel: 208-368-4000

www.micron.com/products/support Sales inquiries: 800-932-4992
Micron and the Micron logo are trademarks of Micron Technology, Inc.
All other trademarks are the property of their respective owners.
This data sheet contains minimum and maximum limits specified over the power supply and temperature range set forth herein.
Although considered final, these specifications are subject to change, as further product development and data characterization some-
times occur.

Automotive DDR4 SDRAM: MT40A512M8 MT40A256M16 Features

Uploaded by

Copyright:

Available Formats

Automotive DDR4 SDRAM: MT40A512M8 MT40A256M16 Features

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Automotive DDR4 SDRAM: MT40A512M8 MT40A256M16 Features

Uploaded by

Copyright:

Available Formats

4Gb: x8, x16 Automotive DDR4 SDRAM

Automotive DDR4 SDRAM

Table 1: Key Timing Parameters

Note: 1. Refer to the Speed Bin Tables for backward compatibility

Parameter 512 Meg x 8 256 Meg x 16

Note: 1. Page size is per bank, calculated as follows:

Figure 1: Order Part Number Example

07$ &RQILJXUDWLRQ 3DFNDJH 6SHHG 5HYLVLRQ

( W&. QV&/ 

WRITE Preamble ........................................................................................................................................ 59

VREFDQ Calibration Enable .......................................................................................................................... 65

Per-DRAM Addressability .............................................................................................................................. 115

Changing Refresh Rate ............................................................................................................................... 149

CRC Data Bit Mapping ............................................................................................................................... 169

READ Burst Operation ............................................................................................................................... 198

Write Timing – Data Strobe-to-Data Relationship ........................................................................................ 223

ODT Read Disable State Table .................................................................................................................... 254

IDD, IPP, and IDDQ Measurement Conditions ................................................................................................ 312

Figure 51: CA Parity Error Checking – SRX Attempt ........................................................................................ 111

Figure 103: CRC Write Data Operation .......................................................................................................... 166

Figure 120: Read Timing Definition ............................................................................................................... 192

Figure 201: ODT During Reads ...................................................................................................................... 257

Figure 239: tAOFAS and tAONAS Definitions .................................................................................................. 306

Figure 240: Thermal Measurement Point ....................................................................................................... 312

Table 51: Power-Down Entry Definitions ....................................................................................................... 156

Important Notes and Warnings

General Notes and Description

x8 configurations. The DDR4 SDRAM uses an 8n-prefetch architecture to achieve high-

• Any specific requirement takes precedence over a general statement.

Definitions of the Device-Pin Signal Level

Definitions of the Bus Signal Level

Functional Block Diagrams

Figure 2: 512 Meg x 8 Functional Block Diagram

577S 577Q 577Z

Figure 3: 256 Meg x 16 Functional Block Diagram

577S 577Q 577Z

Figure 4: 78-Ball x4, x8 Ball Assignments

Notes: 1. See Ball Descriptions.

Figure 5: 96-Ball x16 Ball Assignments

Notes: 1. See Ball Descriptions.

Table 3: Ball Descriptions

Symbol Type Description

Table 3: Ball Descriptions (Continued)

Symbol Type Description

Table 3: Ball Descriptions (Continued)

Symbol Type Description

Figure 6: 78-Ball FBGA – ×4, ×8, "RH"

0.8 TYP 1.1 ±0.1

6.4 CTR 0.34 ±0.05

Notes: 1. All dimensions are in millimeters.

Figure 7: 78-Ball FBGA – ×4, ×8, "SA"

0.8 TYP 1.1 ±0.1

6.4 CTR 0.34 ±0.05

Notes: 1. All dimensions are in millimeters.

Figure 8: 96-Ball FBGA – ×16, "GE"

0.8 TYP 1.1 ±0.1

6.4 CTR 0.29 MIN

Notes: 1. All dimensions are in millimeters.

Figure 9: 96-Ball FBGA – ×16, "LY"

07$ &RQILJXUDWLRQ 3DFNDJH 6SHHG 5HYLVLRQ

( W&. QV&/