COMPUTER ORGANIZATION AND DESIGN

5
Edition
th

The Hardware/Software Interface

Chapter 4
The Processor

4.1 Introduction

Introduction

CPU performance factors

Instruction count

CPI and Cycle time

Determined by CPU hardware, ISA

We will examine two MIPS implementations

Determined by ISA and compiler

A simplified version
A more realistic pipelined version

Simple subset, shows most aspects

Memory reference: lw, sw

Arithmetic/logical: add, sub, and, or, slt
Control transfer: beq, j
Chapter 4 The Processor 2

Instruction Execution Steps

PC (Program Counter) instruction memory,

fetch instruction
Register numbers register file, read registers
Depending on instruction class

Use ALU to calculate

Arithmetic result
Memory address for load/store
Branch target address

Access data memory for load/store

PC target address or PC + 4

Write results to registers register file

Chapter 4 The Processor 3

CPU Overview

Chapter 4 The Processor 4

Multiplexers

Cant just join

wires together

Use multiplexers

Chapter 4 The Processor 5

Control

Op code
Chapter 4 The Processor 6

Information encoded in binary

Combinational element

Low voltage = 0, High voltage = 1

One wire per bit
Multi-bit data encoded on multi-wire buses

4.2 Logic Design Conventions

Logic Design Basics

Operate on data
Output is a function of input

State (sequential) elements

Store information
Chapter 4 The Processor 7

Combinational Elements

AND-gate

Y = A& B

A
B

Multiplexer

A
+

Y = A+ B

Adder

Y = S ? I1 : I0
I0
I1

M
u
x

Arithmetic/Logic Unit

Y = F(A, B)
A
ALU

B
F
Chapter 4 The Processor 8

Sequential Elements

Register: stores data in a circuit

Uses a clock signal to determine when to

update the stored value
Edge-triggered: update when Clk changes
from 0 to 1
Clk

D
Clk

D
Q

Chapter 4 The Processor 9

Sequential Elements

Register with write control

Only updates on clock edge when write

control input is 1
Used when stored value is required later
Clk

D
Write
Clk

Write
D
Q

Chapter 4 The Processor 10

Clocking Methodology

Combinational logic transforms data

during clock cycles

Between clock edges

Input from state elements, output to state
element
Longest delay determines clock period

Chapter 4 The Processor 11

Datapath

Elements that process data and addresses

in the CPU

4.3 Building a Datapath

Building a Datapath

Registers, ALUs, muxs, memories,

We will build a MIPS datapath

incrementally

Refining the overview design

Chapter 4 The Processor 12

Instruction Format
For ALUcontrol

R-type

Load/
Store

Branch

rs

rt

rd

shamt

funct

31:26

25:21

20:16

15:11

10:6

5:0

35 or 43

rs

rt

address

31:26

25:21

20:16

15:0

rs

rt

address

31:26

25:21

20:16

15:0

opcode

always
read

read,
except
for load

write for
R-type
and load

sign-extend
and add

Chapter 4 The Processor 13

Full Datapath

Chapter 4 The Processor 14

Instruction Fetch (for all inst.)

32-bit
register

Increment by
4 for next
instruction

Chapter 4 The Processor 15

R-Format Instructions

Read two register operands

Perform arithmetic/logical operation
Write register result

Chapter 4 The Processor 16

Load/Store Instructions

Read register operands

Calculate address using 16-bit offset

Use ALU, but sign-extend offset

Load: Read memory and update register

Store: Write register value to memory

Chapter 4 The Processor 17

Branch Instructions

Read register operands

Compare operands

Use ALU, subtract and check Zero output

Calculate target address

Sign-extend displacement
Shift left 2 places (word displacement)
Add to PC + 4

Already calculated by instruction fetch

Chapter 4 The Processor 18

Branch Instructions
Just
re-routes
wires

Sign-bit wire
replicated
Chapter 4 The Processor 19

Composing the Elements

First-cut data path does an instruction in

one clock cycle

Share the date path for all instructions

Each datapath element can only do one
function at a time
Hence, we need separate instruction and data
memories; add other elements if needed

Use multiplexers where alternate data

sources are used for different instructions
Chapter 4 The Processor 20

R-Type/Load/Store Datapath

Chapter 4 The Processor 21

Full Datapath

Chapter 4 The Processor 22

ALU can perform multiple functions

Load/Store: F = add
Branch: F = subtract (beq, bne)
R-type: F depends on funct field
ALU control

Function

0000

AND

0001

OR

0010

add

0110

subtract

0111

set-on-less-than

1100

NOR

4.4 A Simple Implementation Scheme

ALU Control

Chapter 4 The Processor 23

ALU Control

Assume 2-bit ALUOp derived from opcode

Combinational logic derives ALU control

Truth Table:

opcode

ALUOp

Operation

funct

ALU function

ALU control

lw

00

load word

XXXXXX

add

0010

sw

00

store word

XXXXXX

add

0010

beq

01

branch equal

XXXXXX

subtract

0110

R-type

10

add

100000

add

0010

subtract

100010

subtract

0110

AND

100100

AND

0000

OR

100101

OR

0001

set-on-less-than

101010

set-on-less-than

0111

Chapter 4 The Processor 24

The Main Control Unit

Control signals derived from instruction

For ALUcontrol

R-type

Load/
Store

Branch

rs

rt

rd

shamt

funct

31:26

25:21

20:16

15:11

10:6

5:0

35 or 43

rs

rt

address

31:26

25:21

20:16

15:0

rs

rt

address

31:26

25:21

20:16

15:0

opcode

always
read

read,
except
for load

write for
R-type
and load

sign-extend
and add

Chapter 4 The Processor 25

Datapath With Control

Shared by
ALL instructions

Chapter 4 The Processor 26

R-Type Instruction

Chapter 4 The Processor 27

Load Instruction

Chapter 4 The Processor 28

Branch-on-Equal Instruction

Chapter 4 The Processor 29

Implementing Jumps
Jump

address

31:26

25:0

Jump uses word address

Update PC with concatenation of

Top 4 bits of old PC

26-bit jump address
00

Need an extra control signal decoded from

opcode
Chapter 4 The Processor 30

Datapath With Jump Added

Chapter 4 The Processor 31

Performance Issues

Longest delay determines clock period

Not feasible to vary period for different

instructions
Violates design principle

Critical path: load instruction

Instruction memory register file ALU
data memory register file

Making the common case fast

We will improve performance by pipelining

Chapter 4 The Processor 32

Pipelined laundry: overlapping execution

Parallelism improves performance

Four loads:

Speedup
= 8/3.5 = 2.3

Non-stop:

4.5 An Overview of Pipelining

Pipelining Analogy

Start up

Speedup
= 2n/0.5n + 1.5 4
= number of stages

Chapter 4 The Processor 33

MIPS Pipeline

Five stages, one step per stage

1.
2.
3.
4.
5.

IF: Instruction fetch from memory

ID: Instruction decode & register read
EX: Execute operation or calculate address
MEM: Access memory operand
WB: Write result back to register

Chapter 4 The Processor 34

Pipeline Performance

Assume time for stages is

100ps for register read or write

200ps for other stages

Compare pipelined datapath with single-cycle

datapath

Instr

Instr fetch Register

read

ALU op

Memory
access

Register
write

Total time

lw

200ps

100 ps

200ps

200ps

100 ps

800ps

sw

200ps

100 ps

200ps

200ps

R-format

200ps

100 ps

200ps

beq

200ps

100 ps

200ps

700ps
100 ps

600ps
500ps

Chapter 4 The Processor 35

Pipeline Performance
Single-cycle (Tc= 800ps)

Pipelined (Tc= 200ps)

Chapter 4 The Processor 36

Pipeline Speedup

If all stages are balanced

i.e., all take the same time

Time between instructionspipelined

= Time between instructionsnonpipelined
Number of stages

If not balanced, speedup is less

Speedup due to increased throughput

Overlap instruction execution

Latency (time for each instruction) does not
decrease
Chapter 4 The Processor 37

Pipelining and ISA Design

MIPS ISA designed for pipelining

All instructions are 32-bits

Few and regular instruction formats

Can decode and read registers in one step

Load/store addressing

Easier to fetch and decode in one cycle

c.f. x86: 1- to 17-byte instructions

Can calculate address in 3rd stage, access memory

in 4th stage

Alignment of memory operands

Memory access takes only one cycle

Chapter 4 The Processor 38

MIPS Pipeline Graphical

5 stage pipeline

200ps cycle time

Read register happened in the 2nd half (100ps)
Write register happened in the 1st half (100ps)
Read can see write if happened in the same cycle

Chapter 4 The Processor 39

Basic RISC Pipelining

Basic idea:

Each instruction spends 1 clock cycle in each of the 5

execution stages.
During 1 clock cycle, the pipeline can be processing
(different stages of) 5 different instructions.

40

Hazards

Situations that prevent starting the next

instruction in the next cycle
Structure hazards

Data hazard

A required resource is busy

Need to wait for previous instruction to
complete its data read/write

Control hazard

Deciding on control action depends on

previous instruction
Chapter 4 The Processor 41

Structure Hazards

Conflict for use of a resource

In MIPS pipeline with a single memory

Load/store requires data access

Instruction fetch would have to stall for that
cycle

Would cause a pipeline bubble

Hence, pipelined datapaths require

separate instruction/data memories

Or separate instruction/data caches

Chapter 4 The Processor 42

Data Hazards

An instruction depends on completion of

data access by a previous instruction

add
sub

$s0, $t0, $t1

$t2, $s0, $t3

Chapter 4 The Processor 43

Forwarding (aka Bypassing)

Use result when it is computed

Dont wait for it to be stored in a register

Requires extra connections in the datapath

Chapter 4 The Processor 44

Load-Use Data Hazard

Cant always avoid stalls by forwarding

If value not computed when needed

Cant forward backward in time!

Chapter 4 The Processor 45

Code Scheduling to Avoid Stalls

Reorder code to avoid use of load result in

the next instruction
C code for A = B + E; C = B + F;

stall

stall

lw
lw
add
sw
lw
add
sw

$t1,
$t2,
$t3,
$t3,
$t4,
$t5,
$t5,

0($t0)
4($t0)
$t1, $t2
12($t0)
8($t0)
$t1, $t4
16($t0)

13 cycles

lw
lw
lw
add
sw
add
sw

$t1,
$t2,
$t4,
$t3,
$t3,
$t5,
$t5,

0($t0)
4($t0)
8($t0)
$t1, $t2
12($t0)
$t1, $t4
16($t0)

11 cycles
Chapter 4 The Processor 46

Control Hazards

Branch determines flow of control

Fetching next instruction depends on branch

outcome
Pipeline cant always fetch correct instruction

Still working on ID stage of branch

In MIPS pipeline

Need to compare registers and compute

target early in the pipeline
Add hardware to do it in ID stage

Chapter 4 The Processor 47

Stall on Branch

Wait until branch outcome determined

before fetching next instruction

Note, assume branch is resolved in 2nd stage!

could be more bubbles, if not!

Chapter 4 The Processor 48

Branch Prediction

Longer pipelines cant readily determine

branch outcome early

Predict outcome of branch

Stall penalty becomes unacceptable

Only stall if prediction is wrong

In MIPS pipeline

Can predict branches not taken

Fetch instruction after branch, with no delay

Chapter 4 The Processor 49

MIPS with Predict Not Taken

Prediction
correct

Prediction
incorrect
Note lw is cancelled!

Chapter 4 The Processor 50

More-Realistic Branch Prediction

Static branch prediction

Based on typical branch behavior

Example: loop and if-statement branches

Predict backward branches taken

Predict forward branches not taken

Dynamic branch prediction

Hardware measures actual branch behavior

e.g., record recent history of each branch

Assume future behavior will continue the trend

When wrong, stall while re-fetching, and update history

Chapter 4 The Processor 51

Pipeline Summary
The BIG Picture

Pipelining improves performance by

increasing instruction throughput

Subject to hazards

Executes multiple instructions in parallel

Each instruction has the same latency
Structure, data, control

Instruction set design affects complexity of

pipeline implementation
Chapter 4 The Processor 52

4.6 Pipelined Datapath and Control

MIPS Pipelined Datapath

MEM

Right-to-left
flow leads to
hazards

WB

Chapter 4 The Processor 53

Pipeline registers

Need registers between stages

To hold information produced in previous cycle

Chapter 4 The Processor 54

Pipeline Operation

Cycle-by-cycle flow of instructions through

the pipelined datapath

Single-clock-cycle pipeline diagram

c.f. multi-clock-cycle diagram

Shows pipeline usage in a single cycle

Highlight resources used
Graph of operation over time

Well look at single-clock-cycle diagrams

for load & store
Chapter 4 The Processor 55

IF for Load, Store,

Chapter 4 The Processor 56

ID for Load, Store,

Chapter 4 The Processor 57

EX for Load

Chapter 4 The Processor 58

MEM for Load

Chapter 4 The Processor 59

WB for Load

Wrong
register
number

Chapter 4 The Processor 60

Corrected Datapath for Load

Chapter 4 The Processor 61

EX for Store

Chapter 4 The Processor 62

MEM for Store

Chapter 4 The Processor 63

WB for Store

Chapter 4 The Processor 64

Multi-Cycle Pipeline Diagram

Form showing resource usage

Chapter 4 The Processor 65

Multi-Cycle Pipeline Diagram

Traditional form

Chapter 4 The Processor 66

Single-Cycle Pipeline Diagram

State of pipeline in a given cycle

Chapter 4 The Processor 67

Pipelined Control (Simplified)

Chapter 4 The Processor 68

Pipelined Control

Control signals derived from instruction

As in single-cycle implementation

Chapter 4 The Processor 69

Pipelined Control

Chapter 4 The Processor 70

Consider this sequence:

sub
and
or
add
sw

$2, $1,$3
$12,$2,$5
$13,$6,$2
$14,$2,$2
$15,100($2)

We can resolve hazards with forwarding

4.7 Data Hazards: Forwarding vs. Stalling

Data Hazards in ALU Instructions

How do we detect when to forward?

Chapter 4 The Processor 71

Dependencies & Forwarding

Chapter 4 The Processor 72

Detecting the Need to Forward

Pass register numbers along pipeline

ALU operand register numbers in EX stage

are given by

e.g., ID/EX.RegisterRs = register number for Rs

sitting in ID/EX pipeline register

ID/EX.RegisterRs, ID/EX.RegisterRt

Data hazards when

1a. EX/MEM.RegisterRd = ID/EX.RegisterRs
1b. EX/MEM.RegisterRd = ID/EX.RegisterRt
2a. MEM/WB.RegisterRd = ID/EX.RegisterRs
2b. MEM/WB.RegisterRd = ID/EX.RegisterRt

Fwd from
EX/MEM
pipeline reg
Fwd from
MEM/WB
pipeline reg

Chapter 4 The Processor 73

Detecting the Need to Forward

1. But only if forwarding instruction will

write to a register!

EX/MEM.RegWrite, MEM/WB.RegWrite

2. And only if Rd for that instruction is not

$zero

EX/MEM.RegisterRd 0,
MEM/WB.RegisterRd 0

Chapter 4 The Processor 74

Forwarding Paths

One forwarding path

Chapter 4 The Processor 75

Forwarding Conditions

EX hazard

if (EX/MEM.RegWrite and (EX/MEM.RegisterRd 0)

and (EX/MEM.RegisterRd = ID/EX.RegisterRs))
ForwardA = 10
if (EX/MEM.RegWrite and (EX/MEM.RegisterRd 0)
and (EX/MEM.RegisterRd = ID/EX.RegisterRt))
ForwardB = 10

MEM hazard

if (MEM/WB.RegWrite and (MEM/WB.RegisterRd 0)

and (MEM/WB.RegisterRd = ID/EX.RegisterRs))
ForwardA = 01
if (MEM/WB.RegWrite and (MEM/WB.RegisterRd 0)
and (MEM/WB.RegisterRd = ID/EX.RegisterRt))
ForwardB = 01
Chapter 4 The Processor 76

Double Data Hazard

Consider the sequence:

add $1,$1,$2
add $1,$1,$3
add $1,$1,$4

Both hazards occur

Want to use the most recent

Revise MEM hazard condition

Only fwd if EX hazard condition isnt true

Chapter 4 The Processor 77

Revised Forwarding Condition

MEM hazard

if (MEM/WB.RegWrite and (MEM/WB.RegisterRd 0)

and not (EX/MEM.RegWrite and (EX/MEM.RegisterRd 0)
and (EX/MEM.RegisterRd = ID/EX.RegisterRs))
and (MEM/WB.RegisterRd = ID/EX.RegisterRs))
ForwardA = 01

if (MEM/WB.RegWrite and (MEM/WB.RegisterRd 0)

and not (EX/MEM.RegWrite and (EX/MEM.RegisterRd 0)
and (EX/MEM.RegisterRd = ID/EX.RegisterRt))
and (MEM/WB.RegisterRd = ID/EX.RegisterRt))
ForwardB = 01

Chapter 4 The Processor 78

Datapath with Forwarding

Chapter 4 The Processor 79

Load-Use Data Hazard

Need to stall
for one cycle

Chapter 4 The Processor 80

Load-Use Hazard Detection

Check when using instruction is decoded

in ID stage
ALU operand register numbers in ID stage
are given by

Load-use hazard when

IF/ID.RegisterRs, IF/ID.RegisterRt
ID/EX.MemRead and
((ID/EX.RegisterRt = IF/ID.RegisterRs) or
(ID/EX.RegisterRt = IF/ID.RegisterRt))

If detected, stall and insert bubble

Chapter 4 The Processor 81

How to Stall the Pipeline

Force control values in ID/EX register

to 0

EX, MEM and WB do nop (no-operation)

Prevent update of PC and IF/ID register

Using instruction is decoded again

Following instruction is fetched again
1-cycle stall allows MEM to read data for lw

Can subsequently forward to EX stage

Chapter 4 The Processor 82

Stall/Bubble in the Pipeline

Stall inserted
here

Chapter 4 The Processor 83

Stall/Bubble in the Pipeline

Or, more
accurately

Chapter 4 The Processor 84

Datapath with Hazard Detection

Chapter 4 The Processor 85

Adding the Hazard Hardware

PCSrc

Hazard
Unit

PC

EX/MEM

Control 0

Register File
Instruction
Memory

Read
Address

ID/EX.MemRead

IF/ID
Add

ID/EX

Shift
left 2

Add

Read Addr 1
Read
1
Read Addr Data
2
Write Addr Read
Data 2
Write Data
16

Sign
Extend

32

ID/EX.RegisterRt

MEM/WB

Branch

Data
Memory
ALU

Address

Read
Data

Write Data
ALU
cntrl

Forward
Unit

Stalls and Performance

The BIG Picture

Stalls reduce performance

But are required to get correct results

Compiler can arrange code to avoid

hazards and stalls

Requires knowledge of the pipeline structure

Chapter 4 The Processor 87

If branch outcome determined in MEM

4.8 Control Hazards

Branch Hazards

If target and outcome is

available on the 4th cycle

Flush these
instructions
(Set control
values to 0)

PC
Chapter 4 The Processor 88

Pipeline
Note, PC updated with branch target address in MEM stage

Chapter 4 The Processor 89

Reducing Branch Delay

Move hardware to determine outcome to ID

stage

Target address adder

Register comparator

Example: branch taken

36:
40:
44:
48:
52:
56:
72:

sub
beq
and
or
add
slt
...
lw

$10,
$1,
$12,
$13,
$14,
$15,

$4,
$3,
$2,
$2,
$4,
$6,

$8
7
$5
$6
$2
$7

# 7*4 + 44 = 72

$4, 50($7)
Chapter 4 The Processor 90

Example: Branch Taken

Resolve branch
in ID stage

Chapter 4 The Processor 91

Example: Branch Taken

Chapter 4 The Processor 92

Data Hazards for Branches

If a comparison register is a destination of

2nd or 3rd preceding ALU instruction

add $1, $2, $3

IF

add $4, $5, $6

beq $1, $4, target

ID

EX

MEM

WB

IF

ID

EX

MEM

WB

IF

ID

EX

MEM

WB

IF

ID

EX

MEM

WB

Can resolve using forwarding

Chapter 4 The Processor 93

Data Hazards for Branches

If a comparison register is a destination of

preceding ALU instruction or 2nd preceding
load instruction

lw

Need 1 stall cycle

$1, addr

IF

add $4, $5, $6

beq stalled
beq $1, $4, target

ID

EX

MEM

WB

IF

ID

EX

MEM

WB

IF

ID
ID

EX

MEM

WB

Chapter 4 The Processor 94

Data Hazards for Branches

If a comparison register is a destination of

immediately preceding load instruction

lw

Need 2 stall cycles (since the source is

needed in ID)
$1, addr

IF

beq stalled
beq stalled
beq $1, $0, target

ID

EX

IF

ID

MEM

WB

ID
ID

EX

MEM

WB

Chapter 4 The Processor 95

Dynamic Branch Prediction

As the amount of ILP exploited increases (CPI

decreases), impact of control stalls increases.

Branches come more often

A longer n-cycle delay postpones more instructions

Dynamic Hardware Branch Prediction

Learns which branches are taken, or not based on

branch history
Make the right guess (most of the time) about whether
a branch is taken, or not.
Delay depends on whether prediction is correct, and
whether branch is taken.
96

Branch-Prediction Buffers (BPB)

Also called branch history table
Low-order n bits of branch address used to index

a table of branch history data for prediction.

May have collisions between distant branches.

Associative tables also possible (expensive)

In each entry, k bits of information about history

of that branch are stored (also called predictor).

Common values of k: 1, 2, and larger

Entry is used to predict what branch will do.

Actual outcome of branch will update the entry.
97

1-bit Branch Predictor

The entry for a branch has only two states:

Bit = 1

The last time this branch was encountered, it was

taken. I predict it will be taken next time.

Bit = 0

The last time this branch was encountered, it was

not taken. I predict it will not be taken next time.

Will make 2 mistakes each time a loop is

encountered.

At the end of the first & last iterations.

May always mispredict in pathological cases!

98

1-Bit Predictor: Shortcoming

Inner loop branches mispredicted twice!

outer:

inner:

beq , , inner

beq , , outer

Mispredict as taken on last iteration of inner

loop
Then mispredict as not taken on first
iteration of inner loop next time around
Chapter 4 The Processor 99

2-Bit Predictor

Only change prediction on two successive

mispredictions (NOTE, this state information is
saved in branch history table for prediction)

Chapter 4 The Processor 100

Branch prediction example

Branch Outcomes: TFTFTTTTFTTTT
1-bit predictor
(Initial state: T)

YNNNNYYYNNYYY

2-bit predictor
(Initial state: TT)

YNYNYYYYNYYYY

Chapter 4 The Processor 101

Implementing Branch Histories

(Important information not in the book!)

Separate cache (prediction table) accessed during
IF stage (to obtain prediction outcome)
Extra bits in instruction cache
Problem with this approach in MIPS:

After fetch, dont know whether the instruction is really a

branch or not (until decoding)
Also dont know the target address.
In MIPS, by the time you know these things (in ID), you
already know whether its really taken!
Havent saved any time!
Branch-Target Buffers (BTB) can fix this problem (later)...

102

Branch Target Buffer

Target address

1-bit predictor
103

Branch Resolution in 5-stage pipeline

Need to resolve branch and
target address in IF stage!!

Fetch from target

This flowchart based on 1-bit predictor

104

Important part of processor design

Unexpected events requiring change
in flow of control

Different ISAs use the terms differently

Exception

Arises within the CPU

e.g., undefined opcode, overflow, syscall,

Interrupt

4.9 Exceptions

Exceptions and Interrupts

From an external I/O controller

Dealing with them without sacrificing

performance is hard
Chapter 4 The Processor 105

Handling Exceptions

In MIPS, exceptions managed by a System

Control Coprocessor (CP0)
Save PC of offending (or interrupted) instruction

In MIPS: Exception Program Counter (EPC)

Save indication of the problem

In MIPS: Use Cause register

Well assume 1-bit

0 for undefined opcode, 1 for overflow

Jump to handler at 8000 00180

Chapter 4 The Processor 106

An Alternate Mechanism

Vectored Interrupts

Example:

Handler address determined by the cause

Undefined opcode:
Overflow:
:

C000 0000
C000 0020
C000 0040

Instructions either

Deal with the interrupt, or

Jump to real handler

Chapter 4 The Processor 107

Handler Actions

Read cause, and transfer to relevant

handler
Determine action required
If restartable

Otherwise

Take corrective action

use EPC to return to program
Terminate program
Report error using EPC, cause,

(like a procedure call to the handler routine)

Chapter 4 The Processor 108

Exceptions in a Pipeline

Another form of control hazard

Consider overflow on add in EX stage
add $1, $2, $1
Prevent $1 from being clobbered (no update)
Complete previous instructions
Flush add and subsequent instructions
Set Cause and EPC register values
Transfer control to handler

Similar to mispredicted branch, i.e. branch to

the handler

Use much of the same hardware

Chapter 4 The Processor 109

Pipeline with Exceptions

Overflow
detected

Select handler

Chapter 4 The Processor 110

Exception Properties

Restartable exceptions

Pipeline can flush the instruction

Handler executes, then returns to the
instruction

Refetched and executed from scratch

PC saved in EPC register

Identifies causing instruction

Actually PC + 4 is saved

Handler must adjust back to PC

Chapter 4 The Processor 111

Exception Example

Exception on add in
40
44
48
4C
50
54

sub
and
or
add
slt
lw

$11,
$12,
$13,
$1,
$15,
$16,

$2, $4
$2, $5
$2, $6
$2, $1
$6, $7
50($7)

Handler (store information, $26, $27 OS use)

80000180
80000184

sw
sw

$26, 1000($0)
$27, 1004($0)

Chapter 4 The Processor 112

Exception Example

Chapter 4 The Processor 113

Exception Example

Chapter 4 The Processor 114

Multiple Exceptions

Pipelining overlaps multiple instructions

Simple approach: deal with exception from

earliest instruction

Could have multiple exceptions at once

Flush subsequent instructions

Precise exceptions

In complex pipelines

Multiple instructions issued per cycle

Out-of-order completion
Maintaining precise exceptions is difficult!

Chapter 4 The Processor 115

Imprecise Exceptions

Just stop pipeline and save state

Including exception cause(s)

Let the handler work out

Which instruction(s) had exceptions

Which to complete or flush

May require manual completion

Simplifies hardware, but more complex handler

software
Not feasible for complex multiple-issue
out-of-order pipelines
Chapter 4 The Processor 116

Correct Exception Handling

In multiple-issue out-of-order pipelines

Instruction issued in order, execution out-oforder

Using reorder buffer to enforce instruction
*commit* in order

Buffer exception, Go down the pipeline

and handle them in commit order

Chapter 4 The Processor 117

Pipelining: executing multiple instructions in

parallel
To increase ILP

Deeper pipeline

Less work per stage shorter clock cycle

Multiple issue

Replicate pipeline stages multiple pipelines

Start multiple instructions per clock cycle
CPI < 1, so use Instructions Per Cycle (IPC)
E.g., 4GHz 4-way multiple-issue

4.10 Parallelism via Instructions

Instruction-Level Parallelism (ILP)

16 BIPS, peak CPI = 0.25, peak IPC = 4

But dependencies reduce this in practice

Chapter 4 The Processor 118

Multiple Issue

Static multiple issue

Compiler groups instructions to be issued together

Packages them into issue slots
Compiler detects and avoids hazards

Dynamic multiple issue

CPU examines instruction stream and chooses

instructions to issue each cycle
Compiler can help by reordering instructions
CPU resolves hazards using advanced techniques at
runtime

Chapter 4 The Processor 119

Speculation Reduced Stalls

Guess what to do with an instruction

Start operation as soon as possible

Check whether guess was right

If so, complete the operation

If not, roll-back and do the right thing

Common to static and dynamic multiple issue

Examples

Speculate on branch outcome

Roll back if path taken is different

Speculate on load

Roll back if location is updated

Chapter 4 The Processor 120

Compiler/Hardware Speculation

Compiler can reorder instructions

e.g., move load before branch

Can include fix-up instructions to recover
from incorrect guess

Hardware can look ahead for instructions

to execute Out-of-order execution

Buffer results until it determines they are

actually needed
Flush buffers on incorrect speculation
Chapter 4 The Processor 121

Speculation and Exceptions

What if exception occurs on a

speculatively executed instruction?

Static speculation

e.g., speculative load before null-pointer

check
Can add ISA support for deferring exceptions

Dynamic speculation

Can buffer exceptions until instruction

completion (which may not occur)

Chapter 4 The Processor 122

Static Multiple Issue

Compiler groups instructions into issue

packets

Group of instructions that can be issued on a

single cycle
Determined by pipeline resources required

Think of an issue packet as a very long

instruction

Specifies multiple concurrent operations

Very Long Instruction Word (VLIW)
Chapter 4 The Processor 123

Scheduling Static Multiple Issue

Compiler must remove some/all hazards

Reorder instructions into issue packets

No dependencies with a packet
Possibly some dependencies between
packets

Varies between ISAs; compiler must know!

Pad with nop if necessary

Chapter 4 The Processor 124

MIPS with Static Dual Issue

Two-issue packets

One ALU/branch instruction

One load/store instruction
64-bit aligned

ALU/branch, then load/store

Pad an unused instruction with nop

Address

Instruction type

Pipeline Stages

ALU/branch

IF

ID

EX

MEM

WB

n+4

Load/store

IF

ID

EX

MEM

WB

n+8

ALU/branch

IF

ID

EX

MEM

WB

n + 12

Load/store

IF

ID

EX

MEM

WB

n + 16

ALU/branch

IF

ID

EX

MEM

WB

n + 20

Load/store

IF

ID

EX

MEM

WB

Chapter 4 The Processor 125

MIPS with Static Dual Issue

Chapter 4 The Processor 126

Hazards in the Dual-Issue MIPS

More instructions executing in parallel

EX data hazard

Forwarding avoided stalls with single-issue

Now cant use ALU result in load/store in same packet

Load-use hazard

add $t0, $s0, $s1

load $s2, 0($t0)
Split into two packets, effectively a stall

Still one cycle use latency, but now two instructions

More aggressive scheduling required

Chapter 4 The Processor 127

Scheduling Example

Schedule this for dual-issue MIPS

Loop: lw
addu
sw
addi
bne
Loop:

$t0,
$t0,
$t0,
$s1,
$s1,

0($s1)
$t0, $s2
0($s1)
$s1,4
$zero, Loop

#
#
#
#
#

$t0=array element
add scalar in $s2
store result
decrement pointer
branch $s1!=0

ALU/branch

Load/store

cycle

nop

lw

addi $s1, $s1,4

nop

addu $t0, $t0, $s2

nop

bne

sw

$s1, $zero, Loop

$t0, 0($s1)

$t0, 4($s1)

IPC = 5/4 = 1.25 (c.f. peak IPC = 2)

Chapter 4 The Processor 128

Loop Unrolling

Replicate loop body to expose more

parallelism

Reduces loop-control overhead

Use different registers per replication

Called register renaming

Avoid loop-carried anti-dependencies

Store followed by a load of the same register

Aka name dependence

Reuse of a register name

Chapter 4 The Processor 129

Loop Unrolling Example

Loop:

ALU/branch

Load/store

cycle

addi $s1, $s1,16

lw

$t0, 0($s1)

nop

lw

$t1, 12($s1)

addu $t0, $t0, $s2

lw

$t2, 8($s1)

addu $t1, $t1, $s2

lw

$t3, 4($s1)

addu $t2, $t2, $s2

sw

$t0, 16($s1)

addu $t3, $t4, $s2

sw

$t1, 12($s1)

nop

sw

$t2, 8($s1)

sw

$t3, 4($s1)

bne

$s1, $zero, Loop

IPC = 14/8 = 1.75

Closer to 2, but at cost of registers and code size

Chapter 4 The Processor 130

Dynamic Multiple Issue

(NOTE, for this course, we do not cover

this advanced topic!)
Superscalar processors
CPU decides whether to issue 0, 1, 2,
each cycle

Avoiding structural and data hazards

Avoids the need for compiler scheduling

Though it may still help

Code semantics ensured by the CPU
Chapter 4 The Processor 131

Dynamic Pipeline Scheduling

Allow the CPU to execute instructions out

of order to avoid stalls

But commit result to registers in order

Example

lw
$t0, 20($s2)
addu $t1, $t0, $t2
sub
$s4, $s4, $t3
slti $t5, $s4, 20
Can start sub while addu is waiting for lw

Chapter 4 The Processor 132

Dynamically Scheduled CPU

Preserves
dependencies

Hold pending
operands

Results also sent

to any waiting
reservation
stations
Reorders buffer for
register writes

Can supply
operands for
issued instructions
Chapter 4 The Processor 133

Register Renaming

Reservation stations and reorder buffer

effectively provide register renaming
On instruction issue to reservation station

If operand is available in register file or reorder

buffer

Copied to reservation station

No longer required in the register; can be
overwritten

If operand is not yet available

It will be provided to the reservation station by a

function unit
Register update may not be required
Chapter 4 The Processor 134

Speculation

Predict branch and continue issuing

Dont commit until branch outcome

determined

Load speculation

Avoid load and cache miss delay

Predict the effective address

Predict loaded value
Load before completing outstanding stores
Bypass stored values to load unit

Dont commit load until speculation cleared

Chapter 4 The Processor 135

Why Do Dynamic Scheduling?

Why not just let the compiler schedule

code?
Not all stalls are predicable

Cant always schedule around branches

e.g., cache misses

Branch outcome is dynamically determined

Different implementations of an ISA have

different latencies and hazards

Chapter 4 The Processor 136

Does Multiple Issue Work?

The BIG Picture

Yes, but not as much as wed like

Programs have real dependencies that limit ILP
Some dependencies are hard to eliminate

Some parallelism is hard to expose

Limited window size during instruction issue

Memory delays and limited bandwidth

e.g., pointer aliasing

Hard to keep pipelines full

Speculation can help if done well

Chapter 4 The Processor 137

Power Efficiency

Complexity of dynamic scheduling and

speculations requires power
Multiple simpler cores may be better

Microprocessor

Year

Clock Rate

Pipeline
Stages

Issue
width

Out-of-order/
Speculation

Cores

Power

i486

1989

25MHz

No

5W

Pentium

1993

66MHz

No

10W

Pentium Pro

1997

200MHz

10

Yes

29W

P4 Willamette

2001

2000MHz

22

Yes

75W

P4 Prescott

2004

3600MHz

31

Yes

103W

Core

2006

2930MHz

14

Yes

75W

UltraSparc III

2003

1950MHz

14

No

90W

UltraSparc T1

2005

1200MHz

No

70W

Chapter 4 The Processor 138

Processor

ARM A8

Intel Core i7 920

Market

Personal Mobile Device

Server, cloud

Thermal design power

2 Watts

130 Watts

Clock rate

1 GHz

2.66 GHz

Cores/Chip

Floating point?

No

Yes

Multiple issue?

Dynamic

Dynamic

Peak instructions/clock cycle

Pipeline stages

14

14

Pipeline schedule

Static in-order

Dynamic out-of-order
with speculation

Branch prediction

2-level

2-level

1st level caches/core

32 KiB I, 32 KiB D

32 KiB I, 32 KiB D

2nd level caches/core

128-1024 KiB

256 KiB

3rd level caches (shared)

2- 8 MB

4.11 Real Stuff: The ARM Cortex-A8 and Intel Core i7 Pipelines

Cortex A8 and Intel i7

Chapter 4 The Processor 139

ARM Cortex-A8 Pipeline

Chapter 4 The Processor 140

ARM Cortex-A8 Performance

Chapter 4 The Processor 141

Core i7 Pipeline

Chapter 4 The Processor 142

Core i7 Performance

Chapter 4 The Processor 143

Unrolled C code

1 #include <x86intrin.h>
2 #define UNROLL (4)
3
4 void dgemm (int n, double* A, double* B, double* C)
5 {
6 for ( int i = 0; i < n; i+=UNROLL*4 )
7
for ( int j = 0; j < n; j++ ) {
8
__m256d c[4];
9
for ( int x = 0; x < UNROLL; x++ )
10
c[x] = _mm256_load_pd(C+i+x*4+j*n);
11
12
for( int k = 0; k < n; k++ )
13
{
14
__m256d b = _mm256_broadcast_sd(B+k+j*n);
15
for (int x = 0; x < UNROLL; x++)
16
c[x] = _mm256_add_pd(c[x],
17
_mm256_mul_pd(_mm256_load_pd(A+n*k+x*4+i), b));
18
}
19
20
for ( int x = 0; x < UNROLL; x++ )
21
_mm256_store_pd(C+i+x*4+j*n, c[x]);
22 }
23 }

4.12 Instruction-Level Parallelism and Matrix Multiply

Matrix Multiply

Chapter 4 The Processor 144

Assembly code:

1 vmovapd (%r11),%ymm4
# Load 4 elements of C into %ymm4
2 mov %rbx,%rax
# register %rax = %rbx
3 xor %ecx,%ecx
# register %ecx = 0
4 vmovapd 0x20(%r11),%ymm3
# Load 4 elements of C into %ymm3
5 vmovapd 0x40(%r11),%ymm2
# Load 4 elements of C into %ymm2
6 vmovapd 0x60(%r11),%ymm1
# Load 4 elements of C into %ymm1
7 vbroadcastsd (%rcx,%r9,1),%ymm0
# Make 4 copies of B element
8 add $0x8,%rcx # register %rcx = %rcx + 8
9 vmulpd (%rax),%ymm0,%ymm5
# Parallel mul %ymm1,4 A elements
10 vaddpd %ymm5,%ymm4,%ymm4
# Parallel add %ymm5, %ymm4
11 vmulpd 0x20(%rax),%ymm0,%ymm5
# Parallel mul %ymm1,4 A elements
12 vaddpd %ymm5,%ymm3,%ymm3
# Parallel add %ymm5, %ymm3
13 vmulpd 0x40(%rax),%ymm0,%ymm5
# Parallel mul %ymm1,4 A elements
14 vmulpd 0x60(%rax),%ymm0,%ymm0
# Parallel mul %ymm1,4 A elements
15 add %r8,%rax
# register %rax = %rax + %r8
16 cmp %r10,%rcx
# compare %r8 to %rax
17 vaddpd %ymm5,%ymm2,%ymm2
# Parallel add %ymm5, %ymm2
18 vaddpd %ymm0,%ymm1,%ymm1
# Parallel add %ymm0, %ymm1
19 jne 68 <dgemm+0x68>
# jump if not %r8 != %rax
20 add $0x1,%esi
# register % esi = % esi + 1
21 vmovapd %ymm4,(%r11)
# Store %ymm4 into 4 C elements
22 vmovapd %ymm3,0x20(%r11)
# Store %ymm3 into 4 C elements
23 vmovapd %ymm2,0x40(%r11)
# Store %ymm2 into 4 C elements
24 vmovapd %ymm1,0x60(%r11)
# Store %ymm1 into 4 C elements

4.12 Instruction-Level Parallelism and Matrix Multiply

Matrix Multiply

Chapter 4 The Processor 145

Performance Impact

Chapter 4 The Processor 146

4.14 Fallacies and Pitfalls

Fallacies

Pipelining is easy (!)

The basic idea is easy

The devil is in the details

e.g., detecting data hazards

Pipelining is independent of technology

So why havent we always done pipelining?

More transistors make more advanced techniques
feasible
Pipeline-related ISA design needs to take account of
technology trends

e.g., predicated instructions

Chapter 4 The Processor 147

Pitfalls

Poor ISA design can make pipelining

harder

e.g., complex instruction sets (VAX, IA-32)

e.g., complex addressing modes

Significant overhead to make pipelining work

IA-32 micro-op approach
Register update side effects, memory indirection

e.g., delayed branches

Advanced pipelines have long delay slots

Chapter 4 The Processor 148

ISA influences design of datapath and control

Datapath and control influence design of ISA
Pipelining improves instruction throughput
using parallelism

4.14 Concluding Remarks

Concluding Remarks

More instructions completed per second

Latency for each instruction not reduced

Hazards: structural, data, control

Multiple issue and dynamic scheduling (ILP)

Dependencies limit achievable parallelism

Complexity leads to the power wall
Chapter 4 The Processor 149