Flow Setup

EDI System Foundation Flow User Guide
Product Version 14.2

October 2014
Document Last Updated: October 2014
2014-2015 Cadence Design Systems, Inc. All rights reserved.

Printed in the United States of America.
Cadence Design Systems, Inc. (Cadence), 2655 Seely Ave., San Jose, CA 95134, USA.
Trademarks: Trademarks and service marks of Cadence Design Systems, Inc. (Cadence)
contained in this document are attributed to Cadence with the appropriate symbol. For queries
regarding Cadences trademarks, contact the corporate legal department at the address shown
above or call 1-800-862-4522.
All other trademarks are the property of their respective holders.
Restricted Print Permission: This publication is protected by copyright and any unauthorized use
of this publication may violate copyright, trademark, and other laws. Except as specified in this
permission statement, this publication may not be copied, reproduced, modified, published,
uploaded, posted, transmitted, or distributed in any way, without prior written permission from
Cadence. This statement grants you permission to print one (1) hard copy of this publication subject
to the following conditions:
1. The publication may be used solely for personal, informational, and noncommercial purposes;
2. The publication may not be modified in any way;
3. Any copy of the publication or portion thereof must include all original copyright, trademark,
and other proprietary notices and this permission statement; and
4. Cadence reserves the right to revoke this authorization at any time, and any such use shall be
discontinued immediately upon written notice from Cadence.
Disclaimer: Information in this publication is subject to change without notice and does not
represent a commitment on the part of Cadence. The information contained herein is the proprietary
and confidential information of Cadence or its licensors, and is supplied subject to, and may be
used only by Cadences customer in accordance with, a written agreement between Cadence and
its customer. Except as may be explicitly set forth in such agreement, Cadence does not make, and
expressly disclaims, any representations or warranties as to the completeness, accuracy or
usefulness of the information contained in this document. Cadence does not warrant that use of
such information will not infringe any third party rights, nor does Cadence assume any liability for
damages or costs of any kind that may result from use of such information.
Restricted Rights: Use, duplication, or disclosure by the Government is subject to restrictions as
set forth in FAR52.227-14 and DFAR252.227-7013 et seq. or its successor.

Table of Contents
Contents
1
About This Manual
11
Overview
11
Introduction to Foundation Flow
11
Foundation Flow Overview Diagram
12
Before You Begin
12
Input for Running Foundation Flows
13
Additional Inputs for Running Foundation Flows
13
Flow Variations
13
About the Flow Documentation
14
15
TCL Variables
15
Installing Foundation Flows
15
Starting Foundation Flows
16
Describing setup.tcl
17
Defining Floorplanning, Scan, CTS, SDC, and Routing Layer
18
Defining Hierarchical Partition Information
19
Defining Library and Technology File
21
Example of Library and Technology File
22
Defining RC Corner Information
22
Example of RC Corner Information
25
Defining Delay Corner Information
25
Example of Delay Corner Information
30
Defining Constraint Modes and Analysis Views
30
Example of Constraint Modes and Analysis Views
31
October 2014

Table of Contents
Defining OpenAccess
33
Mailing Results and Updates
35
Miscellaneous Variables
36
Describing edi_config.tcl
37
Defining Extraction Options
38
Defining Noise Settings
39
Defining Power and Ground Nets
39
Defining Multi-Threading or Distributed Processing
40
Example of Distributed Processing
41
Defining LP.CPF Flow Variables
43
Defining Optional Variables
45
Defining Tie and Filler Cells
46
Defining Welltap and Endcap Specifications
47
Defining Welltap and Endcap Specifications (Per power domain basis)
48
Reporting Power
50
Defining Hier ILMs
51
Defining Flow Control
52
Defining Command Mode Variables
55
Defining ccoptDesign for Clock Tree Construction
58
Defining CCOpt Top and Bottom Layers
59
Defining CCOpt Variables
60
Example Settings for Each Script
61
63
Plug-in Variables and Tags
63
Defining Plug-in Variables
63
Commands that can be Tagged
66
Tags for EDI System Flow
67
Example 1:
67
Example 2:
68
Example 3:
69
October 2014

Table of Contents
70
Using the Wizard
70
Starting the Flow Wizard
71
View and Save Scripts
73
Setting Up the Library for the Flow Wizard
74
Adding Library to the Flow - Summary
75
Setting Up the Library Information
77
Designing the Flow
79
Designing the Flow - Netlist
79
Designing the Flow - Clock Tree
83
Adding Design to the Flow - Summary
84
Timing the Flow
86
Timing the Flow - Timing
86
Timing the Flow - Library Set
88
Timing the Flow - RC Corner
90
Timing the Flow - Delay Corner
92
Timing the Flow - Constraint Mode
94
Timing the Flow - Analysis View
96
Reviewing the Timing Setup
98
Adding Power to the Flow
100
Adding Power to the Flow - Summary
103
Setting up the Tool
104
Setting up the Tool - Setup
105
Setting up the Tool - Placement
106
Setting up the Tool - Optimization
109
Setting up the Tool - CTS
112
Setting up the Tool - Route
114
Setting up the Tool - Signal Intergrity
116
Setting up the Tool - Design For Manufacture (DFM)
119
Setting up the Tool - Multi-CPU
121
October 2014

Table of Contents
Setting up the Tool - Multi-CPU - Distributed Processing
123
Reviewing the Tool Setup
124
Setting up Plug-in Scripts
126
Completing Setup
129
132
Code Generator
132
Introduction to Code Generator
132
Advantages
133
TCL Command
135
Command Line Options
135
Usage
136
Execution Tips
137
Makefile Options - Script Updates
138
Makefile Options - Single Process Execution
138
Example Scripts
139
SINGLE STEP
139
STEP RANGE
139
ALL STEPS
140
HIERARCHICAL FLOW - Example Scripts
141
Results
144
Flat Mode (all steps)
144
Hierarchical Mode (all steps)
145
Error.Warning Messages
148
Warning on Missing Required Data
149
Warning to Replace Old Variables
149
151
CPF-Based Low Power Flow
151
Introduction to CPF-based Low Power Implementation
151
Flow Overview of CPF-based Low Power Flow
152
Before you Begin Implementing CPF-based Low Power Flow
152
October 2014

Table of Contents
Input for CPF-based Low Power Flow
153
Additional Input
154
Defining Variables and Specifying Values for Command Modes
154
Describing lp_config.tcl
155
Timing Environment Initialization
159
MMMC Timing and SI Setup for Low Power P_R Implementation
159
Creating Variables for CPF-based Low Power Flow
162
162
Create Library Sets
162
Create Delay Corners
163
Create Constraint Modes
164
Create Analysis Views
165
Attach RC Corners to Delay Corners
166
Update library_sets with SI Libraries, if defined
166
Define setup and hold Analysis Views and Enable the Default Views
168
Results for CPF-based Low Power Flow
168
Example - CPF-Based Low Power Foundation Flow
169
Plug-in scripts usage in the reference design:
169
CPF File_3a
170
post_cts_tcl_3a
195
post_init_tcl_3a
196
post_place_tcl_3a
200
post_prects_tcl_3a
202
pre_cts_tcl_3a
203
pre_init_tcl_3a
203
pre_place_tcl_3a
203
211
Hierarchical Flow
211
Introduction to Hierarchical Implementation Flow
211
Overview of Hierarchical Implementation Flow
212
October 2014

Table of Contents
One-Pass and Two -Pass Hierarchical Flows
212
Before You Begin - Hierarchical Implementation Flow
212
Inputs for Hierarchical Implementation Flow
213
Hierarchical Partition Flow
214
Implementing Top-Level and Blocks Concurrently
215
Using Interface Logic Models
215
Using the Flattened and Unflattened ILM States
216
Performing Multi-Mode Multi-Corner Timing Analysis and Optimization
217
Managing Clock Latencies
217
Deriving Clock Tree Estimation for Budgeting and CTS
218
Preventing Hierarchical Signal Integrity Issues
222
About the Hierarchical One-Pass ILM Flow and Diagram
222
Hierarchical Foundation Flow with FlexILM
223
Setting Up the Flow
224
Single Pass Hierarchical Flow
225
FlexILM Generation and Top preCTS Optimization with FlexILM
225
Reassigning PTN Pin and Rebudgeting Timing
226
Results for Hierarchical Flow Implementation
226
228
Tags for Flow
228
10
243
Sample Script - Code Generator
243
Sample Single Script Flow
243
Executing the Flow
243
Executing the Flow
October 2014
243

About This Manual
About This Manual

This manual describes the tasks in the recommended Foundation Flow using the Cadence^^
Encounter Digital Implementation System (EDI System) software to implement a block or flat chip
from a completed floorplan, while meeting timing and physical design requirements and resolving
signal integrity (SI) problems.
The EDI System family encompasses the following products:
Encounter Digital Implementation System L
Encounter Digital Implementation System XL
NanoRoute Ultra SoC Routing Solution
Virtuoso Digital Implementation
Encounter Timing System L
Encounter Timing System XL
Encounter Power System L
Encounter Power System XL
First Encounter L
First Encounter XL
First Encounter GXL
For information about these products, see the Getting Started chapter of the EDI System User
Guide About Encounter System Licenses
For more information about the Encounter family of products, see the following documents. You
can access these and other Cadence documents with the Cadence Help documentation system.
EDI System Product Documentation

October 2014

About This Manual
EDI System Product Documentation

Whats New in EDI System
Provides information about new and changed features in this release of the Encounter family
of products.
EDI System Known Problems and Solutions
Describes important Cadence Change Requests (CCRs) for the Encounter family of
products, including solutions for working around known problems.
EDI System User Guide
Describes how to install and configure the EDI System software, and provides strategies for
implementing digital integrated circuits.
EDI System Text Command Reference
Describes the Encounter text commands, including syntax and examples.
EDI System Menu Reference
Provides information specific to the forms and commands available from the EDI System
graphical user interface.
EDI System Database Access Command Reference
Lists the EDI System database access commands and provides a brief description of syntax
and usage.
Mixed Signal Interoperability Guide
Describes the digital mixed-signal flow.
README file
Contains installation, compatibility, and other prerequisite information, including a list of
Cadence Change Requests (CCRs) that were resolved in this release. You can read this file
online at downloads.cadence.com .
October 2014
10

Overview
Overview
In this section we cover a basic introduction to how foundation flows are implemented in EDI System.
Before You Begin
Flow Variations

This guide describes the tasks in the recommended flow using the Encounter Digital Implementation System (EDI System)
software to implement a block or a flat chip from a completed floorplan, while meeting timing and physical design
requirements and resolving signal integrity (SI) problems. The guide also includes Low Power and Hierarchical flows.
For designs with multiple operating modes or corners that require optimization, and for more accurate timing, the flow
supports multi-mode multi-corner (MMMC) timing analysis and on-chip variation (OCV). From EDI System 10.1, the flow
takes advantage of the EDI System software multiple-CPU processing capabilities to accelerate the design process and runs
features in a multi-thread or distributed process-mode.
For the main design tasks the flow uses the EDI System super commands. For example, placeDesign, routeDesign, and
optDesign with as few non-default options as possible.
When the tasks described in this guide are complete, you should have a flat design that is ready for DRC and LVS checks
and other sign-off tasks.
This flow is a recommended starting point for you to work with designs for a block or a flat chip. Ultimately, your design might
have a different final set of commands, owing to your specific customization for your design or technology needs.

October 2014
11

Overview

A completed floorplan is a prerequisite for a flat implementation flow. At the end of each main task you optimize the step and
the task is complete when violations are fixed and the timing is acceptable. The flow is complete and the design is ready for
sign-off tasks when the design meets timing and physical design requirements.
Before You Begin

Make sure the following tasks are complete before starting this flow:
Floorplanning, including the following tasks:
Hard block placement
Region definition
I/O pad placement
Creation of placement and routing blockages
Creating the Clock Tree specification file
Creating the RC scale factors
For more information, see the following sections of the EDI System User Guide
Floorplanning the Design
Creating an Initial Floorplan Using Automatic Floorplan Synthesis
Synthesizing Clock Trees Synthesizing Clock Trees
RC Extraction
October 2014
12

Overview

The foundation flows are driven by a combination of TCL scripts and Makefiles. The command scripts are provided in the
software when you install the tool. The user provides library data and options to customise the flow by means of TCL
variables. The flow expects the library data to be in a file called setup.tcl. A full description of the required and optional
variables is provided later in this guide.
Specify the EDI System configuration file (setup.tcl) with the following information:
Timing libraries (*.lib)
LEF libraries (*.lef)
Timing constraints (*.sdc)
Capacitance table or QRC technology file
SI libraries (*.cdb or *.udn) (this is recommended but not required)
User-defined ILMs
Verilog netlist (*.v)
Floorplan file (*.fp or *.def)
Clock Tree Specification file
Scan Chain information (*.tcl or *.def)
GDS Layer map file (*.gds)
Note: The Flat Implementation Flow assumes that timing and routing constraints are feasible.

In addition, depending on the design and technology, you may need to supply values for the following items:
RC scaling factors
OCV derating factors
Metal fill parameters
Filler cell names
Tie cell names
Welltaps and Endcaps
Clock gating cell names
Spare instance names
JTAG instance names
JTAG rows
LSF queue
Dont Use Cells
Delay Cells
CTS Cells
Flow Variations
Based on the timing model you use, there are three variations of the flat implementation flow:
MMMC: the flow uses the MMMC timing throughout the flow. You may use the MMMC flow in designs that require
October 2014
13

Overview
optimization at multiple operating modes or at multiple corners.

The following two are supported for backward compatibility but are not reccomended:
Default: the flow uses the default timing model until the final Timing Analysis step, when it changes to MMMC timing.
Though the default flow is a non-MMMC flow, it changing to MMMC timing for the final timing analysis step gives the
most accurate sign-off timing.
Postroute MMMC: the flow uses the default timing model for the Preroute portion of the flow and MMMC timing after
Routing. You may use the Postroute MMMC flow in designs that require improved co-relation with third-party tools.
Note: Because MMMC timing incorporates timing information for more modes and corners, the postroute MMMC and MMMC
flows may incur memory and run-time degradation.
This guide includes flow charts for the overall Flat Implementation Flow and for each of the major design tasks, along with
brief explanations. It also includes the following additional information in the Cadence EDI System documentation:

This guide includes flow charts for the overall Flat Implementation Flow and for each of the major design tasks, along with
brief explanations. It also includes the following additional information in the Cadence EDI System documentation:
Command syntax and parameter descriptions in the EDI System Text Command Reference
Task information in the EDI System User Guide
October 2014
14

TCL Variables
TCL Variables
In this section we cover the Tool Command Language (Tcl) variables and descriptions for
foundation flows.
The section covers:

All of the EDI System Foundation Flows rely on the Tcl scripts to keep the flows free from errors and
easy to manage. These Tcl scripts are sourced from the main flow script.
You can download the scripts from the software by using one of the following methods:
On the main EDI System menu, select Tools - Foundation Flow Templates on the main EDI
System menu.
Use the following text command:
writeFlowTemplate [-directory directoryName ]

This will create the following directories:
SCRIPTS (the Foundation Flow scripts)
TEMPLATES (Configuration File templates)
EXAMPLES (working design example and sample configuration files)

October 2014
15

TCL Variables

The Foundation Flow requires a configuration file (setup.tcl) that defines the necessary information
to execute the flow including pointers to design data and command option information. The full list
of variables used to control the flow can be found in the documentation or in the examples in the
TEMPLATES directory in the software.
In addition, some of the variable values can be quickly populated from an existing design in the
following way:
1. Load a design (initial netlist, constraints, floorplan)
2. Run the flow wizard (Flows > Foundation Flow Wizard) OR enter the command:
generateFFSetupFile
3. This will generate the following files: setup.auto.tcl and edi_config.auto.tcl
4. Respectively copy to setup.tcl and edi_config.tcl.
5. (optional) Review and edit the files, if necessary.
Or: Copy the template files from TEMPLATES/EDI and edit them manually.
The following Tcl scripts are described in this guide:
Defines the variables to use in the flow. This script is unique for each design and is the only
script you must edit. It is the source of all flow input, and is a superset of the configuration file
that includes design data, library information, and flow control.
This is the EDI Foundation Flow configuration file. It contains the necessary flow options to
drive the Flat and/or Hierarchical implementation flows. This file is optional and is intended to
support design projects where the setup.tcl is shared between team members and
defines common design data such as library, timing, and technology information and
the edi_config.tcl is a local file that contains flow related information that is unique to
a particular block or run.
Contains the Metal Fill rules.
See also,
October 2014
16

TCL Variables
Defines the variables to use in the flow. This script is unique for each design and is the only script
that is required. It contains variable settings that define the necessary information to drive the
foundation flow; including design data, library information, and flow control.
The following variables are defined in setup.tcl. They are put into an array named vars.
To use them enter:
set vars variable_name value
Note: The vars in the column Variable Name in italics are actually placeholders and can be
substituted with your own values. For example in rc_max,cap_table the value rc_max is a
placeholder and can be a value you want to specify in cap_table.
You can also specify a view_definition_tcl file in place of the MMMC vars. For this, you
must define either of the following:
vars(view_definition_tcl)
OR
vars(library_sets), vars(rc_corners), vars(delay_corners),

vars(constraint_modes) and vars(setup/hold_analysis_view) variables.
Note: The selections mentioned above cannot be mixed.
The following types of variables are covered in this section:
Defining OpenAccess
October 2014
17

TCL Variables
Defining Floorplanning, Scan, CTS, SDC, and

Routing Layer
Variable Name
Value
Type
Usage Description
fp_file
string
Specify the name and location of the EDI floorplan file.

Do not use the variable if specifying the def_files
variable.
def_files
name
Specify the name of the initial DEF file. The fp_file is the
alternate recommended file.
cts_spec
list
This file is optional. If omitted EDI System will use the

SDC to build the clock tree.
max_route_layer
integer
Specify the number of routing layers
process
integer
Specify the process node between min=10, max=250

(nm)
scan_def
name
This is optional, specify the name of the Scan DEF file
dbs_dir
string
Specify the string of the database directory (the default is

DBS)
fp_tcl_file
text
script to build floorplan (overrides fp_file)
fp_tcl_proc
text
TCL procedure to build floorplan (overrides fp_file,

fp_tcl_file)
rpt_dir
string
Specify the string of the report directory (the default is

RPT)
<step>,rpt_dir
string
Specify the report directory for a specific step. If this is not

set, vars(rpt_dir) is used.
script_root
string
Specify the location of SCRIPTS directory (the default is

SCRIPTS)
design
name
Name of the design (required)
October 2014
18

TCL Variables
generate_tracks
boolean
Re-generate tracks as against using what is in the

incoming floorplan. The default is True.
netlist
string
Pointer to netlist file (required)
version
integer
EDI System version (9.1.2, 9.1.3, 10.1.0, ...)
cell_check_early
float
Specify for derate cell library checks
cell_check_late
float
Specify for derate cell library checks
honor_pitch
boolean
An option for generateTracks command (enabled by

vars(generate_tracks)
design_root
name
Name of the root directory of the design (required)
critical_range
float
Specify critical range for TNS optimization
For other setup.tcl variables, see:

For edi_config.tcl variables see Describing edi_config.tcl
For examples see Example Settings for Each Script

Varible Name
October 2014
Value
Type
19
Usage Description

TCL Variables
set vars(placement_based_ptn) 1
float
Specify the option to run placement

based partition flow (feed through
insertion/pin assignment).
Default: 0 (to run route based
partition flow)
float
set vars(insert_feedthrough) 1
Specify the option to run

feedthrough buffer insertion.
Default: 1
float
set vars(abutted_design) 1
Specify the option to support

channelless design.
Default: 0
set
vars(use_proto_net_delay_model)
1
float
Specify the option to use picosecond per micro model for quick
timing estimation.
Default: 0
float
set vars(use_flexmodels) 1
Specify this variable to enable

FlexModel flow to reduce netlist
size.
Default: 0
float
set vars(flexmodel_as_ptn) 1
Specify the option to specify one

FlexModel per partition in a design.
Default: 1
float
set vars(budget_mode) giga_opt
Specify this variable to enable

timing budgeting with
GigaOptvirtual.
Possible values are trial_ipo,
giga_opt and
proto_net_delay_model.
Default: trial_ipo
October 2014
20

TCL Variables

Defining OpenAccess

Variable Name
Value
Type
Usage Description
lef_files
list
Specify the list name of the LEF file
library_sets
list
Specify one or more library set names for slow and fast
processes. The library_sets must be one of the Constraint
Modes and a list member.
enable_ldb
list
Pre-compiles .lib to .ldb.
library_set
,timing
list
Specify the .lib files
library_set ,si
list
Specify the .cdb files (optional)
library_set,aocv
list
Points to the AOCV tables which are required for AOCV

analysis
Also see Example of Library and Technology File

October 2014
21

TCL Variables

Defining OpenAccess

Variable Name
Value Type
set vars(library_sets)
slow fast
set vars(slow,timing)
libs/slow.lib libs/rams_slow.lib
set vars(fast,timing)
libs/fast.lib libs/rams_fast.lib
set vars(slow,si)
libs/slow.cdb
set vars(fast,si)
libs/fast.cdb

Variable Name
Value
Type
Usage Description
rc_corners
list
Specify the RC corner names
rc_corner ,cap_table
list
Specify the cap table per RC corner
rc_corner ,T
list
Specify the temperature
rc_corner ,qx_tech_file
list
Specify the QRC technology file
October 2014
22

TCL Variables
rc_corner ,qx_lib_file
list
Specify the QRC library file
rc_corner ,qx_conf_file
list
Specify the QRC configuration file
rc_corner , xcap_factor
list
Specify the scaling factor for coupling

capacitance (post-route)
rc_corner , scale_factor
float*
Specify scaling factors; valid factors are:

pre_route_res_factor,
pre_route_clk_res_factor
pre_route_cap_factor,
pre_route_clk_cap_factor,
post_route_res_factor,
post_route_clk_res_factor,
post_route_cap_factor,
post_route_clk_cap_factor,
post_route_xcap_factor
Note: *post_route* factors can take a triplet
for low medium high effort scale factors.
rc_corner
,pre_route_res_factor
float
Specify the scaling factor for resistance

(pre-route)
Note: Has replaced variable def_res_factor
**
rc_corner
,pre_route_clk_res_factor
float
Specify the scaling factor for clock

resistance (pre-route)
Note: Has replaced variable
def_clk_res_factor **
rc_corner
,pre_route_cap_factor
float
Specify the capacitance scaling factor (preroute)

Note: Has replaced variable def_cap_factor
**
October 2014
23

TCL Variables
rc_corner
,pre_route_clk_cap_factor
float

capacitance (pre-route)
def_clk_cap_factor **
rc_corner
,post_route_res_factor
float
Specify the scaling factor for resistance

(post-route)
Note: Has replaced variable det_res_factor
**
rc_corner
,post_route_clk_res_factor
float

resistance (post-route)
Note: Has replaced variable det_cap_factor
**
rc_corner
,post_route_cap_factor
float
Specify the capacitance scaling factor

(post-route)
Note: Has replaced variable det_cap_factor
**
rc_corner
,post_route_clk_cap_factor
float

det_clk_cap_factor **
rc_corner
,post_route_clk_res_factor
float
Specify the resistance scaling factor (postroute)

det_clk_res_factor **
rc_corner
,post_route_xcap_factor
float
Specify the scaling factor for coupling

Note: Has replaced variable xcap_factor **
October 2014
24

TCL Variables
Pointer to a file with update_rc_corner

command that sets the scale factors
appropriately for the RC corner.
file
rc_corner ,scale_tcl
Note: ** For backward compatibility, the script retains both the variables.
Also see Example of RC Corner Information
Defining OpenAccess

Variable Name
Value Type
set vars(rc_corners)
worst best
set vars(worst,cap_table)
tech/worst/capTable
set vars(worst,T)
125
set vars(best,cap_table)
tech/best/capTable
set vars(best,T)
October 2014
25

TCL Variables
Variable Name
Value
Type
Usage Description
delay_corners
list
Specify the delay

corner names
delay_corner ,library_set
string
Delay corner library set

name
delay_corner ,rc_corner
string
Delay corner RC
corner name
delay_corner ,derate_factor
float
Specify derating
factors*.
Note: The valid factors
are listed below this
table.
delay_corner ,cell_worst_late
float
Specify the setup for

derating of cells
delay_corner ,cell_worst_early
float
Specify the hold

derating for cells
delay_corner ,cell_best_late
float
Specify the setup for

derating of cells
delay_corner ,cell_best_early
float
Specify the hold

derating for cells
delay_corner ,wire_worst_late
float
Specify the setup

derating for nets
delay_corner ,wire_worst_early
float
Specify the hold

derating for nets
delay_corner ,wire_best_late
float
Specify the setup

derating for nets
delay_corner ,wire_best_early
float
Specify the hold

derating for nets
October 2014
26

TCL Variables
float
delay_corner ,derate_tcl
Allows the user to have

a
delay_corner.derate.tcl
file for each delay
corner instead of
having to set one file
per delay in the
setup.tcl.
Foundation flow also
has an analogous
rc,scale_tcl variable for
RC corners called
rc_corner,scale_tcl
delay_corner, cell_check_early
float
Specify the check early

for cells
delay_corner, cell_check_late
float
Specify the check late

for cells
set
vars(delay_corner,late_opcond_library)
list
Specify the i nternal

library name for the
library in which the late
operating condition is
defined. This is not the
library file name.
set
vars(delay_corner,early_opcond_library)
list
Specify the internal

library name for the
library in which the
early operating
condition is defined.
This is not the library
file name.
set vars(delay_corner,late_opcond)
list
Specify the operating

condition to use for
calculating late arrival
times at a single delay
corner.
October 2014
27

TCL Variables
set vars(delay_corner,early_opcond)
list
Specify the operating

condition to use for
calculating early arrival
times at a single delay
corner.
*Derating Factors
The valid derating factors are described below:
Variable Name
Usage Description
data_cell_late
Specify setup derating of cell
data_cell_early
Specify hold derating of cell
data_net_late
Specify setup derating of net
data_net_early
Specify hold derating of net
clock_cell_late
Specify hold derating of clock cell
clock_cell_early
Specify setup derating of clock cell
clock_net_late
Specify hold derating of clock net
clock_net_early
Specify setup derating of clock net
set_timing_derate
Specify after CTS in case you did not define derating on

initialization
Support for Power Domain - Delay Corner Binding Via

October 2014
28

TCL Variables
Support for Power Domain - Delay Corner Binding Via

Variable Name
Usage Description
List of library sets
set vars(delay_corners)
List of delay corners
set vars(power_domains)
List of power domains to bind
set vars(dc,power_domains)
List of power domains
set
vars(dc,ls,power_domains)
List of power domains. This allows bind to a specify

delay corner AND library set
Support for Power Domain - Per Delay Corner AND Per Library Set
Variable Name
Usage Description
List of library sets
List of delay corners
set vars(power_domains)
List of power domains to bind
set
vars(dc1,pd1,early_library_set)
List of delay corners, power domains, and

early library sets
set
vars(dc1,pd1,late_library_set)

late library sets
set vars(dc2,pd2,library_set)

library sets
Also see Example of Delay Corner Information

October 2014
29

TCL Variables

Defining OpenAccess

Variable Name
Value Type
set vars(worst,pre_route_cap_factor)
1.08
set vars(worst,pre_route_res_factor)
1.12

October 2014
30

TCL Variables

Variable Name
Value
Type
Usage Description
constraints_modes
list
Specify the constraint mode names
constraint_mode
,pre_cts_sdc
list
Specify the Pre CTS SDCs
constraints_mode
,post_cts_sdc
list
Specify the Replacement Post CTS

SDCs (optional)
constraints_mode
,incr_cts_sdc
list
Specify the Incremental Post CTS SDCs

(optional)
analysis_views
list
Specify the analysis view names
analysis_view
,delay_corner
string
Specify the delay corners
analysis_view
,constraint_mode
setup_func
Specify the constraint mode
setup_analysis_views
list
Specify the setup analysis views
hold_analysis_views
list
Specify the hold analysis views
default_setup_view
list
Specify the default setup views
default_hold_view
list
Specify the default hold views
step ,active_setup_views
list
Specify the active setup views
step ,active_hold_views
list
Specify the active hold views
step
,setup_analysis_views
list
Specify the setup analysis views
step
,hold_analysis_views
list
Specify the hold analysis views
See also Example of Constraint Modes and Analysis Views

October 2014
31

TCL Variables

Defining OpenAccess

Variable Name
Value Type
set vars(constraint_modes)
func_mode test_mode
set vars(func_mode,pre_cts_sdc)
sdc/functional.sdc
set vars(test_mode,pre_cts_sdc)
sdc/test_mode.sdc
set vars(func_mode,post_cts_sdc)
sdc/postcts.sdc
slow_worst fast_best
set vars(slow_worst,library_set)
slow
set vars(slow_worst,rc_corner)
worst
set vars(fast_best,library_set)
fast
set vars(fast_best,rc_corner)
best
set vars(slow_worst,clock_cell_early)
0.95
set vars(fast_worst,clock_cell_late)
1.05
October 2014
32

TCL Variables
set vars(analysis_views)
func_slow_worst
func_fast_best
test_slow_worst
test_fast_best
set
vars(func_slow_worst,delay_corner)
slow_worst
set
vars(func_slow_worst,constraint_mode)
func_mode
set vars(func_fast_best,delay_corner)
fast_best
set
vars(func_fast_best,constraint_mode)
func_mode
set
vars(test_slow_worst,delay_corner)
slow_worst
set
vars(test_slow_worst,constraint_mode)
test_mode
set vars(test_fast_best,delay_corner)
fast_best
set
vars(test_fast_best,constraint_mode)
test_mode
set vars(setup_analysis_view)
func_slow_worst
test_slow_worst
set vars(hold_analysis_view)
func_fast_best
test_fast_best
set vars(active_setup_views)
func_slow_worst
set vars(active_hold_views)
func_fast_best
test_fast_best
set vars(default_setup_view)
func_slow_worst
set vars(default_hold_view)
func_fast_best
Defining OpenAccess
October 2014
33

TCL Variables
Defining OpenAccess
Defining OpenAccess
Varible Name
Value
Type
Usage Description
oa_ref_lib
string
Specify the list of OpenAccess reference libraries
oa_abstract_name
name
Specify the list of OpenAccess abstract name(s) for the

libraries
oa_layout_name
string
Specify the list of OpenAccess layout view(s) for the

libraries
dbs_format
string
Specify the format for the intermediate database saves.

Use the value oa for OpenAccess.
oa_design_lib and oa_design_cell are also required in
this situation.
netlist_type
string
Specify the format of the initial database. Valid options

are verilog (default) or oa.
If oa is specified the following three oa_design_* variables
are also required.
oa_design_lib
string
Specify the OpenAccess design library
oa_design_cell
strings
Specify the OpenAccess design cell to read and write
oa_design_view
strings
Specify the OpenAccess design view to import at the start

of the flow.
oa_fp
string
Specify the OpenAccess floorplan support.
When running the flow using OpenAccess libraries there are two ways in which to initalise the flow.
With a netlist and floorplan or with an already initialised OA view.
If the design database needs to interoperable with virtuoso it may be necessary to use the
setOaxMode command. This should be included in a plug script.
October 2014
34

TCL Variables


Variable Name
Value Type
Usage Description
mail ,to
text
Specify the address
mail ,steps
text
Specify the procedure

Defining OpenAccess
October 2014
35

TCL Variables
Variable Name
Value
Type
Usage Description
flat
[off,
partial,
full]
Controls the level of script unrolling
verbose
boolean
Adds comments when true
tags,verbose
boolean
Adds tag comments when true
tags,verbosity_level
[low |
high]
When high adds comments for ALL possible tags

in each flow script
log_dir
name
Directory for log files
plug_dir
string
Directory for plug-ins
skew_buffers
list
Specify list of buffers to use during useful skew
set
vars(enable_flexilm)
boolean
Supports FlexILM for preCTS top-level timing

closure.
Default: ILM based flow

Defining OpenAccess
October 2014
36

TCL Variables
This is an optional configuration file intended to support design projects where the setup.tcl is
shared between team members and defines common design data such as library, timing, and
technology information and the edi_config.tcl is a local file that contains flow related
information that is unique to a particular block or run.
The following variables are put in an array named vars and defined in edi_config.tcl.
To use them enter:
set vars variable_name value

Note: The vars in the column Variable Name in italics are actually placeholders and can be
substituted with your own values. For example in rc_max,cap_table the value rc_max is a
placeholder and can be a value you want to specify in cap_table.
The following types of variables are covered in this section:
Reporting Power
Defining Hier ILMs

October 2014
37

TCL Variables

Variable Name
Value Type
Usage Description
relative_c_thresh
float
Ratio of coupling cap to total cap
total_c_thresh
float
Total cap threshold for coupling
coupling_c_thresh
float
Coupling cap threshold
qrc_config_file
file
Specify configuration files for QRC extraction
qrc_layer_map
file
Specify layer map for QRC extraction
qrc_library
directory
Specify library for QRC extraction
For other edi_config.tcl variables, see

Reporting Power
Defining Hier ILMs
For setup.tcl variables see Describing setup.tcl

October 2014
38

TCL Variables

Variable Name
Value Type
Usage Description
delta_delay_threshold
float
Noise delay threshold

Reporting Power
Defining Hier ILMs

Variable Name
Value Type
Usage Description
power_nets
list
List of power nets
ground_nets
list
List of ground nets

October 2014
39

TCL Variables

Reporting Power
Defining Hier ILMs

The variables can be one of the following: lsf, rsh, local, or custom. Depending on the value of the
distribution, additional variables might be required.
Variable Name
Value
Type
Usage Description
distribute
Specify the distribution style [local, custom, rsh,

lsf]
custom,script
script
Specify the custom grid submission script
distribute , queue
string
Specify the LSF Queue (lsf)
distribute , resource
string
Specify the LSF Resources (lsf)
distribute , args
string
Specify the LSF Args (lsf)
distribute,
host_list
list
Specify the list of hosts (rsh)
distribute_timeout
seconds
Tells how long to wait for a distributed host

before it fails
October 2014
40

TCL Variables
local_cpus
integer
Specify the number of threads
remote_hosts
integer
Specify the number of hosts
cpus_per_remote_host
integer
Specify the number of CPUs per remote host
See also Example of Distributed Processing

Reporting Power
Defining Hier ILMs

Example 1
Variable Name
October 2014
Example Value
41

TCL Variables
set vars(distribute)
custom
set vars(local_cpus)
set vars(remote_hosts)
set vars(custom,script)
/grid/sfi/farm/bin/bsub <run limit>

<project name> <resource string> <queue name>
Example 2
Variable Name
Example Value
set vars(distribute)
lsf
set vars(local_cpus)
set vars(remote_hosts)
set vars(lsf,queue)
nx64
set vars(lsf,resource)
<resource string>
The execution host on which you want to run the job.
set vars(lsf,args)
<run limit> <project name>

October 2014
42

TCL Variables

Variable Name
Value Type
Usage Description
cpf_timing
boolean
Specify the CPF timing.
cpf_file
cpf_file_name
Name of the CPF file
resize_shifter_and_iso_insts
boolean
Allow level shifter and resizing

of isolation cell
opconds
list
Specify conditionals list
opcond ,voltage
voltage
Specify voltage
opcond ,temperature
temp
Specify temperature
opcond ,process
process scale
factor
Specify process
opcond ,library_file
file
Specify library file
cpf_isolation
boolean
Specify CPF isolation cells
cpf_keep_rows
boolean
Specify CPF rows to keep
cpf_level_shifter
boolean
Specify CPF level shifters
cpf_power_domain
boolean
Specify CPF power domains
cpf_power_switch
boolean
Specify CPF power switches
cpf_state_retention
boolean
Specify CPF state retentions
Foundation Flow supports the following scenarios:

Using the variable cpf_file displays the results in FF/view_definition.tcl
1. Variables vars(cpf_file)+ vars(cpf_timing) will set the init_cpf_file but
not the init_mmmc_file. The variable init_design will be executed which means that the
timing will be derived from CPF.
loadCPF will be executed.

commitCPF will be executed.
October 2014
43

TCL Variables
The tcl file view_definition.tcl will be created to define RC Corners and to update the
Delay Corners.
2. Variables vars(cpf_file) + !vars(cpf_timing) will not set
the init_cpf_file but will set the init_mmmc_file. The variable init_design will
be executed which means that the timing will be derived from mmmc_file.
loadCPF will be executed.

commitCPF will be executed.
3. No CPF file. init_cpf_file will not be set but init_mmmc_file will be set
and init_design will be executed which means that timing will be derived from mmmc_file.
So, when cpf_file is set and cpf_timing is NOT defined, we default to (1) (assuming the
CPF has full timing). The user must specify vars(cpf_timing) false to get flow (2)
Reporting Power
Defining Hier ILMs

October 2014
44

TCL Variables

Variable Name
Value
Type
Usage Description
use_list
list
List of cells to use regardless of whether they are set

dont use in the library
dont_use_list
list
Specify the list of dont use cells
size_only_file
list
Specify instances that can be resized only during

optimization
dont_use_file
file
File of setDontUse commands created by the user
clock_gating_cells
list
Specify the list of clock gating cells
jtag_cells
list
Specify the list of JTAG instances
cts_cells
list
Specify the CTS buffer cells
jtag_rows
integer
Number of rows for JTAG placement
spare_cells
list
Specify the list of spare cells instances
filler_cells
list
Specify the list of filler cell names
delay_cells
list
Specify the list of delay cell names
tie_cells
list
Specify the list of tie cells [tiehi tielo]
gds_files
list
Specify the GDS file name
gds_layer_map
file
Specify GDSII to LEF layer mapping file
buffer_name
list
Specify the list of assigned buffers.
buffer_tie_assign
boolean
Specify buffer tie high/low to assign nets
assign_buffer
boolean
Add buffers for assign statements. The default is

False.
October 2014
45

TCL Variables

Reporting Power
Defining Hier ILMs

Variable Name
Value
Type
Usage Description
tie_cells
list
Specify the list of tie cells
filler_cells
list
Specify the list of filler cells
tie_cells,
max_distance
vars
Specify the distance between tie-cell and tie-pins less than

given value. A float value (microns).
tie_cells,
max_fanout
vars
Specify the number of tie-pins a tie-net can drive. Zero

means no-limit (integer)

October 2014
46

TCL Variables

Reporting Power
Defining Hier ILMs

The max_gap and cell_interval parameters are mutually exclusive, the user has to define any one
of these parameters to add welltap cells.
Variable Name
Value
Type
Usage Description
welltaps
name
Specify the well tap cells
welltaps ,checkerboard
boolean
Specify true or false in the checkerboard

in well tap cells.
welltaps ,max_gap
float
Specify the maximum distance from the right

edge of one well-tap cell to the left edge of
the following well-tap cell in the same row
welltaps
,cell_interval
float
Specify the maximum distance from the

center
of one well-tap cell to the center of the
following
well-tap cell in the same row
welltaps ,verify_rule
float
Specify the verify rules between welltap cells
October 2014
47

TCL Variables
pre_endcap
cell_name
Specify the Pre Endcap cells
post_endcap
cell_name
Specify the Post Endcap cells

Reporting Power
Defining Hier ILMs
Defining Welltap and Endcap Specifications (Per

power domain basis)
Note: This is optional.
Variable Name
October 2014
Value
Type
Usage Description
48

TCL Variables
domain
,welltaps
name
Specify the well tap cells in the domain
domain
,checkerboard
boolean
Specify true or false in the checkerboard in well tap cells
domain
,max_gap
float
Specify the maximum distance from the right edge of one

well-tap cell to the left edge of the following well-tap cell in
the same row
domain
,cell_interval
float
Specify the maximum distance from the center of one

well-tap cell to the center of the following well-tap cell in
the same row
domain
,pre_endcap
cell_name
Specify the Pre Endcap cells
domain
,post_endcap
cell_name
Specify the Post Endcap cells

Reporting Power
Defining Hier ILMs
October 2014
49

TCL Variables
Reporting Power
Reporting Power
Variable Name
Value
Type
Usage Description
activity_file
string
Specify the name of the activity file
activity_file_format
enum
Specify the format of the activity file TCF | SAF |

VCD
report_power
boolean
Choose True or False depending upon whether

you want power reported
power_analysis_view
list
Specifies the default setup view

Defining Hier ILMs
October 2014
50

TCL Variables
Defining Hier ILMs

Variable Name
Value
Type
Usage Description
ilm_non_sdc_file
ilm
Allows support for interactive Non-SDC constraint

support in ILMs
ilm_list
ilm
Specifies ILMs for bottom up hierarchical design
lef_file
ilm
ilm_dir
ilm
pre_cts_ilm_sdc
mode
enable_dlm
true |
false
Enables new flex ILM hierarchical flow

Reporting Power
October 2014
51

TCL Variables

These variables are optional and are specific to the design.
Variable Name
Value
Type
Usage Description
activity_file
string
Specify the name of the activity file
activity_file_format
enum
Select the format of the activity type

TCF | SAF | VCD
report_power
boolean
Choose True or False depending upon

whether you want power reported
power_analysis_view
list
Specifies the default setup view
enable_ocv
enum
Specify when to enable OCV (list of

enum values). The options are: {false |
pre_postcts | pre_postroute |
pre_postroute_si | pre_prects |
pre_signoff}
enable_cppr
enum
Specify during enable CPPR. The

options are: {setup, hold, both, none}
enable_ss
enum
Specify if/when to enable signalStorm

delay calculation. The options are:
{pre_place | pre_prects | pre_postcts, |
pre_postroute | pre_postroute_si |
pre_signoff | false}
enable_si_aware
boolean
Specify whether to enable si-aware

postroute optimization
Note: This will disable the postroute_si
and postroute_si_hold steps as the are
unnecessary when this is enabled)
October 2014
52

TCL Variables
fix_hold
enum
Controls the points of the flow where

hold optimization is enabled. The
options are: false | postcts | postroute |
postroute_si
catch_errors
string
Specify to save an immediate database

on errors in catch
save_on_catch
string
Saves the EDI database when the catch

is activated because of an error
antenna_diode
string
Cell for antenna fixing
save_constraints
string
Save constraints along with design

database
abort
boolean
Abort the codegen if an error is found .

the default is True.
enable_pac
boolean
Enable post-assembly closure flow. The

default is False.
flow
MMMC
Not necessary in 10.1 onwards

because the flow is always MMMC
report_run_time
boolean
Specify to track and report CPU time for

each step
useful_skew
boolean
Specify to enable useful skew

optimization
postroute_spread_wires
boolean
Enable postroute wire spreading
signoff_extraction_effort
string
Specify the signoff extraction effort
postroute_extraction_effort
enum
Specify level of extraction effort. The

default is medium. This variable is a
flow control variable related to the type
of extraction post route.
Enables the new two pass hierarchical
flow. The valid types are 1pass
(default) and 2pass
hier_flow_type
October 2014
53

TCL Variables
enable_aocv
Enables the new AOCV timing feature
rc_corner,atf_file
A CCOpt technology file (atf) needed for

the new command ccoptDesign
delay_corner,power_domains
Defines the binding between delay

corner and power domains. It is used
with the vars(cpf_timing)
false flow.
absolute_lib_path
Saves absolute paths to library files and

is an option to saveDesign
relative_path
Saves relative paths to data files and is

is an option to saveDesign
cts_engine
Switches on ccopt features. The valid

values are [cts, ccopt, and ccopt_cts]
enable_celtic_steps
setAnalysisMode -aocv
true |
false
Enable the Celtic SI fixing steps

(Default: Off in 12.0)

Reporting Power
Defining Hier ILMs
October 2014
54

TCL Variables


In this section we provide a list of recommendations for the Command Mode variables. The
placement variables assume that all I/Os are placed in legal locations, as they are not moved by
default.
The set optimization options should be persistent throughout the flow. Preserve the assertions on
hierarchical ports (enable for hierarchical flows so that module ports that are constrained are not
replicated).
Note: Values in italic text are recommended but are not default values.
For congested designs, set the congestion effort to high instead of to medium.
Variable Name
Default
place_io_pins
false
clock_gate_aware
true
congestion_effort
auto
in_place_opt
false
no_pre_place_opt
false
place_opt_design
false
For timing effort:

Variable Name
Default
high_timing_effort
false
In hierarchical flow, the final full flat timing analysis is done with the original constraints and it
does not take into account the block level constraint files (that may have been modified). Use
-preserveAssertions {true | false} to avoid issues related to buffer removal. It may be
beneficial to run the first pass of hold fixing with setup degradation disallowed to see how
many hold violations are left, and then check if the violations are real and decide how to
October 2014
55

TCL Variables
proceed.
To enable leakage power optimization throughout the flow, use leakagePowerEffort
{low | high}. An high effort may impact timing and will have an impact on run-time.
For pre-route optimization, set critical range to the percentage of sub-critical paths to be
optimized, for example, 0.2 or 0.5.
Variable Name
Default
preserve_assertions
false
leakage_power_effort
none
dynamic_power_effort
none
fix_hold
true
fix_hold_ignore_ios
false
fix_hold_allow_tns_degradation
true
The high effort hold fixing should improve the quality of results at the expense of runtime.
Set CTS options; the recommendation is to route the clock nets during CTS. If some signals
should not be routed, set the clock nets option to false and use RouteClkNet true in the
spec file.
Variable Name
Default
Usage Description
route_clock_nets
true
Route all clock nets during CTS
clock_eco
false
Run ckECO [postcts, postroute, both, none]
clock_gate_clone
false
Choose True or False depending upon whether you want

to clone the gates
Set the routing options. The recommended options for multicut via insertion and litho driven
routing incur run-time penalties but yield advantages so they are included here. The
routeWithLithoDriven option is recommended for 45 nm but may also be used at
65 nm.
October 2014
56

TCL Variables
The -drouteUseMultiCutViaEffort option can be set to high if needed; expect an

average of 5 percent coverage improvement. Expect an average 10 percent run-time
degradation as compared to the medium effort. Setting the value to high also affects TNS.
Consider disabling this option during early trials.
The flow default is to disable multi-cut via insertion. For 90nm and below, it is recommended
to insert multi-cut vias using set vars(multi_cut_effort) medium. This will come
with a runtime and memory penalty and will also affect QoR. The set
vars(multi_cut_effort) can be set to high if needed; expect an average of 5

percent coverage improvement with an average 10 percent runtime hit versus medium and
will also affect TNS.
Variable Name
Default
Usage Description
multi_cut_effort
low
Insert multi-cut vias during routing
litho_driven_routing
false
Litho-driven routing option
in_route_opt
false
Performs timing optimization after track assignment
Set the noise analysis options. Existing CeltIC users should preserve existing settings if
applicable. Consult the SI Analysis application note for more information on these settings.
You can get the application note from your Cadence Support Representative.
Variable Name
Default
Usage Description
delta_delay_threshold
undefined
Specify the delay noise threshold
celtic_settings
undefined
Specify the Celtic settings
total_c_thresh
undefined
Specify the total cap threshold filter
relative_c_thresh
undefined
Specify the relative threshold filter
coupling_c_thresh
undefined
Specify the coupling threshold filter
si_analysis_type
enum
Specify the analysis type (default or

pessimistic)
acceptable_wns
float
Specify to define the target slack for SI

optimization
October 2014
57

TCL Variables

Reporting Power
Defining Hier ILMs

The following vars are used for defining ccoptDesign for Clock Tree Construction.
Variable Name
Value Type
Usage Description
cts_inverter_cells
list
Specify the CTS inverter cells
cts_buffer_cells
list
Specify the CTS buffer cells
clock_gate_cells
list
Specify the clock gate cells
cts_use_inverters
boolean
Specify true or false
update_io_latency
boolean
Specify true or false
October 2014
58

TCL Variables
cts_target_skew
float
Specify the skew
cts_target_slew
float
Specify the slew
cts_io_opt
enum
Specify on | off | secondary
cts_effort
enum
Specify low | medium | high

The following vars define top and bottom layers for clock tree and leaf nets, non default rules, and
clock shielding net (ccopt only).
Variable Name
Value Type
Usage Description
clk_tree_top_layer
string
Specify setCTSMode | setCCoptMode
clk_tree_bottom_layer
string
clk_leaf_top_layer
string
clk_leaf_bottom_layer
string
clk_tree_ndr
string
clk_leaf_ndr
string
clk_tree_shield_net
string

October 2014
59

TCL Variables

The following variables are added to Foundation Flow to enable Clock Current
Optimization. CCOpt runs clock tree synthesis (CTS) concurrently with placement and physical
optimization and uses a timing window driven engine to optimize timing paths and clocks
simultaneously.
Variable Name
Description
cts_engine
Enables Clock Tree Synthesis, Clock Current Optimization

and Clock Current Optimization enabled CTS (cts, ccopt,
and ccopt_cts).
ccopt_executable
Enables the path to the CCOpt executable. This is optional.
cts_buffer_cells
List of usable buffers (required).
cts_inverter_cells
List of usable inverters (required).
cts_use_inverters
Enables the use of inverters (true or false).
cts_target_slew
Enables maximum clock tree transition time
cts_target_skew
Enables initial skew target.
cts_io_opt
Enables IO timing optimization (on, off, secondary).
cts_effort
Defines the effort level (low, medium, high).
pre_ccopt_tcl
Enables the native pre-ccopt plug-in.
post_ccopt_tcl
Enables the native post-ccopt plug-in.
clk_tree_top_layer
Enables the top tree preferred layer.
clk_tree_bottom_layer
Enables the bottom tree preferred layer.
clk_tree_ndr
Enables the NDR for tree nets.
clk_leaf_top_layer
Enables the top leaf preferred layer.
clk_leaf_bottom_layer
Enables the bottom leaf preferred layer
cts_leaf_ndr
Enables the NDR for leaf nets.
October 2014
60

TCL Variables
clk_tree_shield_net
Enables the shielding net (ccopt only).
clock_gate_cells
This is an existing variable for CTS but is also used for

ccoptDesign
update_io_latency
This is an existing variable for CTS but is also used for

ccoptDesign

Each step of the script has settings as shown in the example below which tells the user which
variables affect the script generation.
Place
set vars(step) place
These variables affect this step:
vars(congestion_effort)
vars(clock_gate_aware)
vars(place_io_pins)
vars(in_place_opt)
vars(preserve_assertions)
vars(leakage_power_effort)
vars(dynamic_power_effort)
vars(critical_range)
PreCTS
set vars(step) prects

October 2014
vars(process)
vars(preserve_assertions)
vars(leakage_power_effort)
vars(dynamic_power_effort)
vars(critical_range)
vars(useful_skew)
61

TCL Variables
- vars(skew_buffers)
CTS
set vars(step) cts

vars(process)
vars(route_clock_nets)
vars(litho_driven_routing)
vars(multi_cut_effort)
PostCTS
set vars(step) postcts

- vars(process)
- vars(enable_cppr)
- vars(clock_gate_clone)
See Sample Script - Code Generator for an example implementation.
October 2014
62


Tags and plug-ins provide an easy method to both expand upon and customise the basic flows. In
this section we describe the pre-defined variables for plug-ins and tags. The section covers:

Plug-in variables are included to allow you to customize the flow. For each step in the flow, there
are plug-ins before and after the main step command. For example, in the placement step, there is a
plug-in before and after the placeDesign command. The plug-ins are called pre_ step _tcl
and post_ step _tcl.
So, in the case of the placement step, the plug-in would be
called vars(pre_place_tcl) and vars(post_place_tcl) .
These variables point to a user generated TCL file containing EDI System commands which are run
in addition to the basic flow commands. The following plug-ins are supported.
Keep in mind that there is no error checking but errors that occur will be caught and the tool will exit
gracefully and save databases, where appropriate.
Note: The plug-in script should not include the restoreDesign command. The foundation flow
automatically restores the design from the previous step.
Plug-in Variable Name
Value
Type
Usage Description
always_source_tcl
string
The contents of the file will be sourced at the start of

each stage of the flow
pre_init_tcl
string
The content of the file will be sourced before loading

the design
October 2014
63

post_init_tcl
string
The content of the file will be sourced after loading the

design
pre_place_tcl
string

design but before running the placement step. Before
running the command placeDesign
place_tcl
string
When specified, this file will replace the default

placement commands
post_place_tcl
string
The content of the file will be sourced after running the

placement step but before exiting the script. After
running the command placeDesign
pre_prects_tcl
string

the CTS
post_prects_tcl
string
The content of the file will be sourced after pre-CTS

optimization.
cts_tcl
string
When specified, this file will replace the default

clockDesign command
pre_cts_tcl
string
The content of the file will be sourced after placing the

design but before running designing the clock. Before
running the command clockDesign
post_cts_tcl
string
The content of the file will be sourced after designing

the clock. After running the command clockDesign
Specify the directory PLUG/post_cts.tcl
pre_postcts_tcl
string
The content of the file will be sourced before post-CTS

optimization
post_postcts_tcl
string
The content of the file will be sourced after post-CTS

optimization
pre_postcts_hold_tcl
string
The content of the file will be sourced before post-CTS

Hold Fixing
post_postcts_hold_tcl
string
The content of the file will be sourced after before postCTS Hold Fixing
October 2014
64

pre_route_tcl
string

CTS but before routing. Before running the command
routeDesign
post_route_tcl
string
The content of the file will be sourced after routing.

After running the command routeDesign
pre_postroute_tcl
string
The content of the file will be sourced before postRoute

Hold Fixing
post_postroute_tcl
string
The content of the file will be sourced after postRoute

Hold Fixing
pre_postroute_hold_tcl
string
The content of the file will be sourced before postRoute

Hold Fixing
post_postroute_hold_tcl
string
The content of the file will be sourced after postRoute

Hold Fixing
pre_postroute_si_tcl
string
The content of the file will be sourced before fixing SI
post_postroute_si_tcl
string
The content of the file will be sourced after fixing SI
pre_postroute_si_hold_tcl
string

the post routing SI hold design
post_postroute_si_hold_tcl
string
The content of the file will be sourced after SI Hold

Fixing
postroute_spread_wires
boolean

post routing design and spreading the wires
pre_signoff_tcl
string
The content of the file will be sourced before final static

timing analysis (STA)
post_signoff_tcl
string
The content of the file will be sourced after before final

static timing analysis (STA)
metal_fill_tcl
string
An example is provided in the TEMPLATES directory
final_always_source_tcl
string
Specify the path to plug-in file. Runs after every flow.
October 2014
65

metalfill_tcl
string
When specified, this file will replace the default metal

fill commands
metalfill
string
Specify variable to define if and when to insertion

metalfill (pre_postroute, pre_postroute_si, pre_signoff)
pre_model_gen_tcl
string
The content of the file will be sourced before running

FlexModel model generation
pre_assign_pin_tcl
string
The content of the file will sourced before running

partition pin assignment
Note: Use the plug-in pre_postroute_si_tcl for noiseTWFmode. To do so, enter:

setSIMode --noiseTwfMode mode
Plug-in Usage:
To enable a plug-in, define the appropriate plug-in variable in the setup.tcl file. The software
sources the plug-in automatically at the appropriate time in the flow. For example:
set vars(post_cts_tcl) PLUG/post_cts.tcl

To use the post_cts_tcl plug-in to adjust clock latencies after CTS.
Below is a list of commands that can be tagged with the following tags:
-pre_tcl filename => Script to be run before the tagged command

-post_tcl filename => Script to be run after the tagged command
-replace_tcl filename => Script to be run to replace the tagged command
-skip true | false => Skip the tagged command
These variables are structure step,commmand,tag
They should be added to your setup.tcl or edi_config.tcl
See Also,
October 2014
66


Tags provides a more granular approach to flow customization for customers who need it. Each
step of the flow has a list of tags and they are generated by gen_setup.tcl in a file called
edi_tags.auto.tcl. Each command has four tags that you need to define. The types of tags are:
A variable to skip the command entirely
A script to completely replace the command
Scripts to run before the command
Scripts to run after the command
The following list lists the tags for the EDI System foundation flow Tags for Flow
This section also contains these examples:
Example 1:
Example 2:
Example 3:
Example 1:
set vars(prects,opt_design,post_tcl) file
set var(flat) full
Where file contains:

saveDesign $vars(dbs_dir)/prects1.enc
setClockDomains -fromType register -toType register
optDesign -preCTS incr
...
optDesign -preCTS -outDir RPT -prefix prects
# <begin tag prects,opt_design,post_tcl>
October 2014
67

saveDesign $vars(dbs_dir)/prects1.enc
setClockDomains -fromType register -toType register
optDesign -preCTS incr
# <end tag prects,opt_design,post_tcl>
...
By setting, vars(flat) to full, the contents of the tag script get imported into the run script to
create a fully flattened script. Otherwise, there would simply be source file.
Example 2:
By default, time_design for place, cts, and route are skipped. If you want to enable timing
reports for any of them, just set vars( step ,time_design,skip) true.
set vars(cts,time_design,skip)
true
...
setCTSMode -routeClkNet true
setNanoRouteMode -routeWithLithoDriven true
createClockTreeSpec
specifyClockTree -update "AutoCTSRootPin * RouteClkNet YES"
clockDesign -skipTimeDesign -outDir RPT
...
# <begin tag cts,time_design,skip>
October 2014
68

# timeDesign -postcts -prefix cts -outDir RPT
# <end tag cts,time_design,skip>
Example 3:
Steps to create directories inside the RPT directory where reports can be appended while running
the flow:
Create a final_always_source_tcl plug-in:
set vars(final_always_source_tcl) file

Where file:
Creates the directory you want.
For Example:
file mkdir $vars(rpt_dir)/$vars(step)

Moves all relevant reports into this directory
For Example:
foreach file glob $vars(rpt_dir) $vars step*

file move $file $vars(rpt_dir) $vars(step)
Add a final timeDesign --reportOnly
For Example:
timeDesign --reportOnly --prefix $vars(step) --outDir

$vars(rpt_dir)/$vars(step)
This has the advantage of being able to easily change the prefix as well.
See also, Plug-in Variables and Tags
October 2014
69

Using the Wizard
Using the Wizard

EDI System provides a foundation flow wizard and code generator to enable you to quickly generate
scripts to load design data, setup timing environment, and execute the recommended implementation
flow from Placement through Signoff.
The EDI System wizard allows you to define all information necessary to generate a setup.tcl file to
drive the foundation flows. This GUI walks the customer through library definition in such a way the
required information can be obtained while hiding the underlying complexities of the MMMC timing
infrastructure. It also provides the user all the necessary foundation flow variable settings. The user can
also re-enter the wizard and restore the current settings.
Note: You can supplement the generated setup.tcl with additional variables from the documentation or
TEMPLATE directory. The wizard does not split the variables
between setup.tcl in edi_config.tcl or support the Low Power and Hierarchical flows
The Wizard can be enabled from the last tab on the tool menu.
Note: All *fields in the wizard are mandatory. All fields without * are optional. The Wizard facilitates your
enteries and shows you a summary at the end of each stage. The files you need for the wizard to run
are placed in the folder ../FF and its sub folders. The system adds the variables you do not enter.
In this section we discuss the following:
In this section, we discuss:
Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup
October 2014
70

Using the Wizard

Change to the directory ff_workshop/wizard and enter this command to start the software:
encounter
This command will start the Encounter Digital Implementation System in the graphical mode.
Choose Flow - Foundation Flow Wizard.
Note: The wizard allows you to start from scratch, from the current design in memory, or from an existing
setup.tcl. In this section we will discuss starting from scratch.
October 2014
71

Using the Wizard
Welcome to Wizard - Basic Fields and Options

How Do You want to Start
October 2014
72

Using the Wizard
Start from
Scratch
Choose this option to automatically configure the Wizard. You will still need to select
files and enter names of files but the background process will take care of most of the
configurations. In this section we will discuss starting from scratch.
Load the
Design
Setup from
Memory
Choose this option if you have already created a setup.tcl file or the system has it in
memory. Click Load.
Load
Previously
Installed
Script
Click the icon on the field to select a file from the system and to modify it as you go
along creating the configurations to save in setup.tcl.
Foundation Flow Install Directory
Select the install directory by browsing the system and click

Install.
Save Foundation Flow Database

At
Select where you want to save the database.
Save Foundation Flow Reports At
Select where you want to save the database.
Continue
Click to move to next screen
For other Wizard forms, click link below:

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup

As you complete your setup you can now View your Foundation Flow scripts. Click the View button on
top right to view the script:
October 2014
73

Using the Wizard
You can also save the script by clicking the Save button top right. Choose where you want to save the
different scripts: Setup, EDI Configure and LP Configure.
Choose the folder and click Apply.

In this section we cover:
October 2014
74

Using the Wizard

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup

The Library summary display.
October 2014
75

Using the Wizard
Setup Library Summary - Basic Fields and Options

Specify Process
Node
Displays the type of node you specified. Click Edit to go back to the Library form
and change value.
Technology LEF
Displays the LEF file you specified. Click Edit to go back to the Library form and
change value.
October 2014
76

Using the Wizard
Physical LEF
Displays the LEF file you specified. Click Edit to go back to the Library form and
change value.
See also,

October 2014
77

Using the Wizard

The following screen appears:
Setting Up the Wizard - Basic Fields and Options

October 2014
78

Using the Wizard
Setting Up the Wizard - Basic Fields and Options

Setting Up the Process Node of Your
Design
Sets up the process node. This is pre-populated,

change it to the value you want.
Specify the Technology LEF File for your

Design
Select the LEF file directory by browsing the system.
Use the File Browser to select Additional

LEF Libraries for your File
Select additional LEF libraries, if any. See Notes
Back
Click to move to the previous screen and change any

entry, if needed.
Continue
Click to move to next screen.
See also,
Designing the Flow

In this section we cover the following
Starting the Flow Wizard Starting the Flow Wizard
Setting Up the Library for the Flow Wizard Setting Up the Library for the Flow Wizard
Designing the Flow Designing the Flow
Timing the Flow
Setting up the Tool Setting up the Tool
Setting up Plug-in Scripts Setting up Plug-in Scripts
Completing Setup Completing Setup

October 2014
79

Using the Wizard
Setup Your Netlist and Floorplan - Basic Fields and Options

October 2014
80

Using the Wizard
Setup Your Netlist and Floorplan - Basic Fields and Options

Choose the Verilog
Netlist for your Design
Select the Verilog netlist

by browsing the system
Design Name
Name of the design.
Yes: Choose the

Floorplan File
Select the Floorplan file by browsing the

system and additionally, Use the DEF file
No
Select to use a Script to Generate the

Complete Floorplan in User Plug-in
Choose the Floorplan

File for your Design
Specify Power and Ground Net Name

Power Net Name
Enter value of power net name.
Ground Net Name
Enter value of power net name
Back
Click to move to the previous screen and change any entry, if needed.
Continue
Notes
October 2014
81

Using the Wizard
Notes
During the course of this wizard you may need to find files from your system. When you click ... the
following type of screen will show up.
Double Pane Selection Form - Fields and Options

Right
Pane
Displays the relevant files and folders in the system. You can browse and select the other
directories.
Add
Select the file that you need and click Add. The right pane opens. Select the file and click
<<. The file is added to the left pane.
Delete
Select the file that you need and click Delete.
Filter
Specifies a filter pattern for selecting the files.
See also,
October 2014
82

Using the Wizard


October 2014
83

Using the Wizard
Setup Your Clock Tree Synthesis Constraints - Basic Fields and Options
Automatically by EDI Using the
Following Clock Cells
Allows the wizard to generate the clock tree

automatically.
Use my Clock Tree Spec File
Select to use your own clock tree specification file from

the system. See Notes
By My User Defined Script
Browse and select your own script
Back

entry, if needed.
Continue
Reviewing the Clock Tree Setup

In the next screen you can review the configurations you have entered. You can edit them or, if
satisfied, click Continue.
See also,
Designing the Flow - Netlist Designing the Flow - Netlist

The Design summary display.
October 2014
84

Using the Wizard
Setup Design Summary - Basic Fields and Options

Netlist
Displays the netlist you specified. Click Edit to go back to the Library form and
change value.
Floorplan
Displays the floorplan you specified. Click Edit to go back to the Library form and
change value.
October 2014
85

Using the Wizard
Clock
Tree
Displays the clock tree you specified. Click Edit to go back to the Library form and
change value.
See also,
Timing the Flow

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup

October 2014
86

Using the Wizard
Timing the Flow - Timing - Basic Fields and Options

October 2014
87

Using the Wizard
Timing the Flow - Timing - Basic Fields and Options

Are you Timing Libraries ECSM or
CCS Model?
Select Yes or No
Do you want to Enable SignalStorm

Delay Calculation?
Select Yes or No. If you select Yes then choose at what

stage from the drop down menu.
Do you want to enable

onChipVariation?
Select Yes or No. If you select Yes then choose at what

stage from the drop down menu.
Back
Click to move to the previous screen and change any entry,

if needed.
Continue
See also,

October 2014
88

Using the Wizard
Create Timing Library Sets - Basic Fields and Options

October 2014
89

Using the Wizard
Create Timing Library Sets - Basic Fields and Options

Create Library
Sets Using the
Tree Browser
Use right mouse click to Add and once added, to Edit or Delete. The following
form opens File Menu > Add Library Set. This form will take you to further forms
where you make selections and then come back to the Wizard.
Back
Continue
See also,

October 2014
90

Using the Wizard
Create RC Extraction Corners - Basic Fields and Options

October 2014
91

Using the Wizard
Create RC Extraction Corners - Basic Fields and Options

Specify the RC
Corner Using
the Tree
Browser
Use right mouse click to Add and once added, to Edit or Delete. The following
form opens File Menu > Add RC Corner. This form will take you to further forms
Back
Continue
See also,

October 2014
92

Using the Wizard
Create Delay Corner Sets - Basic Fields and Options

Do you want to Enable Clock Path Pessimism Removal?
Enable CPPR
After
October 2014
Select the stage from the drop-down menu
93

Using the Wizard
Specify Delay
Corner(s) Using
the Matrix Table
Select the button to turn it On or OFF. When you click OFF the following form
opens File Menu > Add Delay Corner. This form will take you to further forms
Back
Continue
See also,

October 2014
94

Using the Wizard
Create Constraint Modes - Basic Fields and Options

October 2014
95

Using the Wizard
Create Constraint Modes - Basic Fields and Options

Display and Edit the
Constraint Modes Using
the Tree Browser
Use mouse right-click Add Constraint Mode and the form will take you
to File Menu > Add Constraint Mode where you make selections and
then come back to the Wizard.
Back
Continue
See also,

October 2014
96

Using the Wizard
Create Analysis Views - Basic Fields and Options

October 2014
97

Using the Wizard
Create Analysis Views - Basic Fields and Options

Specify Analysis View(s) using the
Matrix Table
Select the button to turn it On or OFF
Specify Default Setup Analysis Views
Specify the view you want. Choose from the drop-down

list
Specify Default Hold Analysis Views
Specify the view you want. Choose from the drop-down

list
Back

entry, if needed.
Continue
See also,

satisfied, click Continue. Depending on your entries, the summary will look like this.
October 2014
98

Using the Wizard
See also,
October 2014
99

Using the Wizard


October 2014
100

Using the Wizard
Setup Your Power - Basic Fields and Options

October 2014
101

Using the Wizard
Setup Your Power - Basic Fields and Options

Specify Default Power
Analysis View
The name of the file appears in a drop-down list, choose it or any other.
Do You Want to
Optimize Power
Select Leakage Power or Dynamic Power to use if you want to optimize

power. Each of the options comes in three flavours: High, Medium, and
Low. This is optional.
Do you have an Activity

File?
Specify the activity file if you have one and select it from the system. Else,
select No. (Default)
Do you want to
Generate Power
Reports?
Select Yes if you want to generate reports. Else, select No. (Default)
Does your Power have

MSO or PSV
Requirements?
Specify the CPF file if you have MSO or PSV requirements and you can
additionally keep DEF rows. Else, select No. (Default)
Do you want to Commit

Certain Portion of the
CPF File?
Select from options if you want to commit certain sections of the CPF
files. This option is active if you specify a CPF file in the previous option.
Back
Continue
Reviewing the Power Setup

satisfied, click Continue.
Designing the Flow
Timing the Flow
Setting up the Tool
October 2014
102

Using the Wizard

Completing Setup

The Power summary display.
October 2014
103

Using the Wizard
Setup Power Summary - Basic Fields and Options

Power Analysis View
Displays the name you specified for the power analysis view.
Optimize Power
Lists your selections.
Activity File
Report Power
CPF File
Edit
Click to go back to the previous screen and make changes.
Setting up the Tool

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup

October 2014
104

Using the Wizard

The following screen appears :
Setup Tool Specific Options Basic Fields and Option

October 2014
105

Using the Wizard
Setup Tool Specific Options Basic Fields and Option

Do you want to Skip the
Tool Setup
Yes Select if you want to skip. No Select if you want to overwrite

the default tool options
Back
Click to move to the previous screen and change any entry, if

needed.
Continue
See also,

If you choose to set up the tool, the following screen appears:
October 2014
106

Using the Wizard
Set up Placement Options - Basic Fields and Option

October 2014
107

Using the Wizard
Set up Placement Options - Basic Fields and Option

Do Not Run Pre-Place
Optimization
Check the box to not run Pre-Placement optimization
Run In-place Optimization
Check the box to run in-place optimization
Place IO Pins
Check the box to place IO pins
Enable Clock Gate Aware If

You Have Clock Gating
Cells in Your Design
Check the box to enable clock gate aware if you have clock gating
cells in your design
Specify Congestion Effort
Select from drop-down list.
Specify Number Of Rows

For JTAG Placement
Enter value to specify number of rows for JTAG placement
Specify Tie Hi/Lo Cells
Select the file from a new screen that shows you the selected Hi/Lo
Cells cells in right hand folder. You can Add or Delete cells to the left
hand folder. Then click Apply
Specify Filler Cells
Select the file from a new screen that shows you the selected Filler
Cells cells in right hand folder. You can Add or Delete cells to the left
Specify Well Taps
Select the file from a new screen that shows you the selected Well
Tap cells in right hand folder. You can Add or Delete cells to the left
Specify JTAG Cells
Select the file from a new screen that shows you the selected JTAG
Cells in right hand folder. You can Add or Delete cells to the left
Select
Click to select the required file from a new screen that shows you the
selected cells in right hand folder. You can Add or Delete cells to the
left hand folder. Then click Apply.
See also,
October 2014
108

Using the Wizard


October 2014
109

Using the Wizard
Set up Optimization Options - Basic Fields and Option

October 2014
110

Using the Wizard
Set up Optimization Options - Basic Fields and Option

Assign Buffer
Check box to assign buffer
Preserve Constraints on IO Port
Check box to preserve constraints on IO ports
Specify critical range of TNS Optimization
Enter value for critical range of TNS optimization
High Timing Effort
Check box for High Timing effort
Enable Clock Gate Aware If You Have Clock

Gating Cells in Your Design
Check box to enable clock gating aware if you have

clock gating cells in your design
Enable Clock Gating
Check box to enable clock gating
Enable useful Skew and Select Skew

Buffers
Select value from drop-down to enable useful skew

and select skew buffers
Enable Hold Fixing
Check box to enable hold fixing
Disable Hold Fixing for IO
Check box to disable hold fixing for IO
Enable Hold Fixing allow TNS degradation
Check box to enable hold fixing allowing TNS

degradation
Resize Shifter and Isolation Instances
Check box to resize shifter and isolation instances
Specify Dont Use List
Enter value for dont use list
Specify Use List
Enter value for use list
Back
Click to move to the previous screen and change

any entry, if needed.
Continue
See also,
October 2014
111

Using the Wizard


October 2014
112

Using the Wizard
Set up CTS Options - Basic Fields and Option

October 2014
113

Using the Wizard
Set up CTS Options - Basic Fields and Option

Disable Routing Clock Nets During
CTS
Check box to disable routing
Enable Clock ECO
Select Post-CTS or Post-Route, or both
Back
To move to the previous screen and change any entry, if

needed.
Continue
See also,

October 2014
114

Using the Wizard
Set up NanoRoute Options - Basic Fields and Option

October 2014
115

Using the Wizard
Set up NanoRoute Options - Basic Fields and Option

Generate Tracks
Check box to generate tracks
Enable Litho Driven Routing
Check box to enable litho driven routing
Enable Wire Spreading at Post-Route

Optimization
Check box to enable wire spreading at post-route

optimization
Specify Top Routing Layer
Select options from drop-down
Set Concurrent via Optimization Effort
Select Low, Medium, or High
Set Post-Route Effort
Select Low, Medium, or High
Back
To move to the previous screen and change any entry,

if needed.
Continue
See also,

October 2014
116

Using the Wizard
Set Up SI Options - Basic Fields and Option

October 2014
117

Using the Wizard
Set Up SI Options - Basic Fields and Option

How Do You Want to Set Up
Extraction Thresholds?
Select radio button if you want to use system defaults or if

you want to customize the thresholds, enter,
Coupling Capacitance Threshold
Relative Capacitance Threshold
Total Capacitance Threshold |
How Do You want to Set Noise

Thresholds?
Select radio button if you want to use system defaults or if

you want to customize the thresholds, enter,
Delta Delay Threshold
Celtic Settings |
How Will You Be Running Signoff

Timing?
Select use EDI/ETS or 3rd party Tool
What Effort Level of Extraction Do

You want for the Signoff Step?
Select High, Medium, Low effort
Back
To move to the previous screen and change any entry, if

needed.
Continue
See also,

October 2014
118

Using the Wizard

Set Up DFM Options - Basic Fields and Option

October 2014
119

Using the Wizard
Set Up DFM Options - Basic Fields and Option

Insert Metal Fill Before
Select where you want to insert metalfill from drop-down menu
Specify Custom Tcl to

Add Metalfill
Select the file from a Windows Explorer format which shows the files in
directories
Open Access Abstract

View Name
directories
Open Access Layout

View Name
directories
Open Access Referance

Library
directories
GDS Map File
directories
List of GDS Files
List of GDS Files opens. Click >> to open right pane for you to browse
the exact file, select and Add.
The file and path shows up on the left pane. Provide a name for the GDS
File and click Close.
Back
Continue
See also,

October 2014
120

Using the Wizard

Tool Setup - Multi-CPU Options - Basic Fields and Option

October 2014
121

Using the Wizard
Tool Setup - Multi-CPU Options - Basic Fields and Option

Number of Local
CPUs
Enter value for number of local CPUs
Number of Remote
Machines
Enter value for number of remote machines
Number of CPUs per

Remote Machine
Enter value for number of CPUs per remote machine
Setup Distribute
Processing
Click to setup distribued processing. the following screen opens: Setting

up the Tool - Multi-CPU - Distributed Processing
Back
Continue
See also,
Setting up the Tool - Multi-CPU - Distributed

October 2014
122

Using the Wizard
Setting up the Tool - Multi-CPU - Distributed

Processing
Tool Setup - Multi-CPU Options - Distributed Processing - Basic Fields and

Option
LSF
Select and enter values for Queue, Resource, Arguments
Custom
Select and enter script path
Local
Select if CPUs are local
October 2014
123

Using the Wizard
RSH
Select and provide Host Name
See also,

satisfied, click Continue. Depending on your entries, the summary will look like this.
October 2014
124

Using the Wizard
Set up Summary Options - Basic Fields and Option

October 2014
125

Using the Wizard
Set up Summary Options - Basic Fields and Option

Place
Lists your Power selections
Optimize
Lists your Optimization selections
CTS
Lists your CTS selections
Route
Lists your Route selections
SI
Lists your SI selections
DFM
Lists your DFM selections
Multi-CPU
Lists your multi-CPU selections
Edit
Click to go back to screen and make changes
See also,

October 2014
126

Using the Wizard
Setup User Plug-in Scripts - Basic Fields and Options

October 2014
127

Using the Wizard
Setup User Plug-in Scripts - Basic Fields and Options

Do you want to Skip User Plug-in?
Yes. No Specify entries if you want to run

your own scripts at specific stages of the
flow.
Always Source at Beginning of Each Step
Check box and select file
Initialization - Before Installation or After Installation or
Placement - Before Placement or During Placement

or After Placement
Pre-CTS Optimization - Before Pre-CTS Optimization

or After pre-CTS Optimization
Clock Tree Synthesis - Before CTS or During CTS or

After CTS
Post CTS Optimization - Before post-CTS

Optimization or After post-CTS Optimization
Routing - Before Routing or After Routing
Post-Route Optimization - Before Post-Route

Optimization or After Post-Route Optimization
Post-Route Optimization + Hold Fixing - Before PostRoute Optimization or After Post-Route Optimization
Post-Route Optimization + Hold Fixing + SI - Before

Post-Route Optimization or After Post-Route
Optimization
Post-Route Signoff - Before Post-Route Signoff or

After Post-Route Signoff
Sign-off - Before Signoff or After Signoff
Back
Click to move to the previous screen and

change any entry, if needed.
Continue
October 2014
128

Using the Wizard

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup
Completing Setup
October 2014
129

Using the Wizard
Completing - Basic Fields and Options

The final screen shows the wizard is now set up and can save entries in a setup.tcl file.
October 2014
130

Using the Wizard

Designing the Flow
Timing the Flow
Setting up the Tool
Completing Setup
October 2014
131

Code Generator
Code Generator
From this release, EDI system provides a code generator to enable users to quickly generate scripts
to load their design data, setup their timing environment, and execute the recommended
implementation flow from placement through signoff.
Note: Foundation Flow code generator generates EDI System 11.1 based scripts automatically. No
changes to setup.tcl are required. With this release, the design import command
init_design replaces the command loadConfig.
Usage
Example Scripts
Results

From this release, EDI system provides a code generator to enable users to quickly generate scripts
to load their design data, setup their timing environment, and execute the recommended
implementation flow from placement through signoff. The feature helps the users overcome the
earlier difficult to read and debug scripts. This feature helps those customers who would like to use
the scripts as a reference flow as well as for auditing purposes where readability is important.
The following diagram depicts how the Code Generator fits into the Foundation Flow.
October 2014
132

Code Generator
The foundation flow is in the SCRIPTS directory and the code generator is
SCRIPTS/gen_flow.tcl (symbolic link to gen_edi_flow.tcl).
To learn more on EDI System Code Generator follow these links:
Advantages
TCL Command
Advantages
The following are the advantages of the Code Generator:
Through the Code Generator the scripts will be generated instead of being executed.
All flow will have the command sequence based on the user settings. A clean, flat script will
be generated for conditional code that is based on static user settings (for example,
setup.tcl).
All conditional code branches that require run time specific data will be generated as
conditionals (that is, they cannot be flattened).
For example, since analysis view can be altered via user plug-ins, checks must be made to
determine which corners are active:
October 2014
133

Code Generator
set vars(active_rc_corners) [list]

foreach view [concat [all_setup_analysis_views] [all_hold_analysis_views]] {
set corner [get_delay_corner [get_analysis_view $view -delay_corner] -rc_corner]
if {[lsearch $vars(active_rc_corners) $corner] == -1 } {
lappend vars(active_rc_corners) $corner
}
The following procedures will be covered in procs.tcl.

Procedure
Type
system_info
Description
Reports system info of the host machine
vars(report_system_info)
source_file
file
Wrapper around source to catch syntax errors
source_plug
plug_in
Loads one or more plug-in scripts, catch errors, etc.
report_time
Reported elapsed CPU time vars(report_run_time)
The code generator is backwards compatible (it will work with existing code).
For the LP flow, there are these additional procedures:
Procedure
Description
get_power_domains
Returns a list of power domains
modify_power_domains
Modifies power domains*
add_power_switches
Adds power switches*
route_secondary_pg_nets
Routes secondary power/ground nets*
insert_welltaps_endcaps
Inserts welltaps and endcaps cells
Note: * operates based on user variable settings in the lp_config.tcl file

These procedures are used by the foundation flow at various points in the scripts. They are also
available for the user to call in plug-in scripts. To access use the following syntax:
FF_EDI::proc_name [option]
Note: Once the scripts have been generated, the use model will be identical to previous releases
When the scripts are generated, the code generator will check for data consistency; i.e. all
October 2014
134

Code Generator
required variables are properly defined, necessary files exists, etc.
TCL Command
Usage: tclsh SCRIPTS/gen_flow.tcl options steps
Step
Description
Execute step a only
a-b
Execute all steps between step a and b, inclusive
a-
Execute all steps in the flow from step a, inclusive
all
Execute all steps in the flow
Note: "a" and "b" should be valid steps for the specific mode.

The following are the command line options:
Command
Description
-h | -help
Print this message
-m | -mode
Set the script mode. Default is "flat".

Valid modes are: flat and hier
-f | -flat
Flatten/unroll levels
- full
Unroll plug-ins and tags (include contents in the main flow script)
- none
Reference plug-ins and tags as procedures.
-d | -dir
The code generator creates a design specific set of run scripts in a FF directory in
the local user directory. You can optionally change the name of the output
directory using this option.
October 2014
135

Code Generator
-s | -setup
Provide the directory containing the setup.tcl setup file required to run the
Foundation Flow.
Default is ..
-e | -edi
Use EDI System to run the code generator vs tclsh. This is necessary if the
setup.tcl utilizes EDI System commands that are not available in tclsh (for
example, getenv)
See also,
Usage
Example Scripts
Results
Usage
Below are the steps to use the Code Generator:
1. Dump the foundation flow code generator from the EDI System interface (GUI or TUI) or
download from the download manager.
GUI: Flows > Create Foundation Flow Template
TUI: writeFlowTemplate --directory directory
2. Create a setup.tcl using either the foundation flow wizard Using the Wizard or using
gen_setup.tcl.
To create using wizard, simply invoke the wizard and follow the on screen instructions.
When finished, save as setup.tcl.
To create using gen_setup.tcl,
load an existing EDI database
and source directory /SCRIPTS/gen_setup.tcl (the existing data
base should be the starting point for the foundation flow)
Edit the resulting setup.auto.tcl and edi_config.auto.tcl files (if
necessary)
October 2014
136

Code Generator
Rename the files to setup.tcl and edi_config.tcl, respectively.

3. Run the code generator: tclsh directory /SCRIPTS/gen_flow.tcl options
4. Execute the flow: make
Note: See the Makefile Options [ on page 129|Makefile Options#886976] for more information.
To learn more on EDI System Code Generator Usage see:
Execution Tips
Execution Tips
Once the flow scripts are generated the use model for the flow execution differs from earlier in these
ways:
The generated scripts are expaned (unrolled) based on the variables in the setup.tcl, so
any change in the setup.tcl will only affect the scripts if they are regenerated.
-
make init
The code generator runs again automatically. This is intended to keep the generated scripts
synchronized with setup.tcl. This can be disabled by running:
make UPDATE=no
Makefile generation can be disabled entirely using the -n option to gen_edi_flow.tcl

make prects
The flow will run the steps from <last completed> to prects
The flow can be executed either in a single process or in multi-process; the default is multiprocess (i.e. each step runs a separate EDI System process). This is more deterministic but
also increases runtime due to re-loading and re-timing the step databases. To run a single
process flow, use
make single FF_STOP=prects
October 2014
137

Code Generator
Any completed step in the flow can be checked for logical equivalence (against
vars(netlist)) by running:
make lec_<step>
Any completed step in the flow loaded into an EDI GUI session by running:
make debug_<step>

Since the script are now dynamic (generated) vs static, a change in the configuration files say
setup.tcl will only take effect when scripts are re-generated. So, the default behavior of the Makefile
is to update the scripts automatically anytime one of the input configuration files change. This
behavior can be disabled, by:
make UPDATE=no
Note: A new option ?u|-rundir allows the scripts/Makefile to be generated elsewhere than
the current work directory ".".
.

The foundation flow defaults to executing each step using different EDI process (a new EDI session
is invoked for each step). To invoke the foundation flow using a single process, use:
make single
Additionally, to invoke a selected set of steps using a single process, use:
make single FF_STOP=end_step
This will start at the current step and stop at the step FF_STOP.
Note: To force the FF to re-run previously executed steps, use FF_START=begin_step or simply
remove the make semaphore files for the steps you want to re-run. The make protected variable or
abstract data type (semaphore) is the step file in the make directory (i.e. make/place, make/prects)
which denotes which steps have completed successfully.
See also,
October 2014
138

Code Generator
Example Scripts
Results
Example Scripts
Following are the examples of the code generation:
SINGLE STEP
STEP RANGE
ALL STEPS
Usage
Results
SINGLE STEP
% tclsh SCRIPTS/gen_flow.tcl init
<FF> Finished loading setup.tcl
<FF> GENERATING STEP init
-------------------------------------------------
<FF> CODE GENERATION COMPLETE
-------------------------------------------------
STEP RANGE
% tclsh SCRIPTS/gen_flow.tcl route-signoff
October 2014
139

Code Generator
<FF> GENERATING STEP route
<FF> GENERATING STEP postroute
<FF> GENERATING STEP postroute_hold
<FF> GENERATING STEP postroute_si_hold
<FF> GENERATING STEP postroute_si
<FF> GENERATING STEP signoff
-------------------------------------------------
-------------------------------------------------
ALL STEPS
w/OUTPUT DIRECTORY RENAMED (-d MYFF)
% tclsh SCRIPTS/gen_flow.tcl -d MYFF all
<FF> OUTPUT DIRECTORY == MYFF
-------------------------------------------------
<FF> Generating scripts for dtmf_recvr_core
-------------------------------------------------
October 2014
140

Code Generator
w/setup.tcl located in another directory (-s MYSETUPDIR) and plug-ins/tags imported (-f full)
% tclsh SCRIPTS/gen_flow.tcl -s MYSETUPDIR -f full all

HIERARCHICAL FLOW
% tclsh SCRIPTS/gen_flow.tcl -m hier al
-------------------------------------------------
-------------------------------------------------
<FF> GENERATING STEP assemble
<FF> Finished loading EDI configuration file
<FF> GENERATING STEP partition
-------------------------------------------------
-------------------------------------------------
<FF> GENERATING STEP place
<FF> GENERATING STEP prects
October 2014
141

Code Generator
<FF> GENERATING STEP cts
<FF> GENERATING STEP postcts
<FF> GENERATING STEP postcts_hold
-------------------------------------------------
<FF> GENERATING flat Makefile ...
-------------------------------------------------
<FF> GENERATING STEP debug
-------------------------------------------------
<FF> Generating scripts for tdsp_core
-------------------------------------------------
<FF> GENERATING STEP place
October 2014
142

Code Generator
<FF> GENERATING STEP prects
<FF> GENERATING STEP cts
<FF> GENERATING STEP postcts
<FF> GENERATING STEP postcts_hold
-------------------------------------------------
<FF> GENERATING flat Makefile ...
-------------------------------------------------
-------------------------------------------------
<FF> GENERATING hier Makefile ...
-------------------------------------------------
October 2014
143

Code Generator
-------------------------------------------------
-------------------------------------------------
See Sample Script - Code Generator for an example implementation.
Results
The code generator default to creating an instance of the foundation flow in a FF directory in the
local user directory. The user can optionally override this using the option
-d directory.
Here is an example of the files created by the code generator:

Usage
Example Scripts
Results

Flat Mode (all steps):
FF
FF/vars.tcl
FF/run.conf
FF/procs.tcl
October 2014
144

Code Generator
FF/EDI
FF/EDI/run_all.tcl
FF/EDI/run_init.tcl
FF/EDI/run_place.tcl
FF/EDI/run_prects.tcl
FF/EDI/run_cts.tcl
FF/EDI/run_postcts.tcl
FF/EDI/run_route.tcl
FF/EDI/run_postroute.tcl
FF/EDI/run_postroute_si.tcl
FF/EDI/run_signoff.tcl
FF/EDI/run_debug.tcl
FF/EDI/gen_html.tcl
Makefile

Hierarchical Mode (all steps):
Makefile
October 2014
145

Code Generator
FF/vars.tcl
FF/run.conf
FF/procs.tcl
FF/EDI/run_assemble.tcl
FF/EDI/run_partition.tcl
FF/EDI/run_debug.tcl
PARTITION
PARTITION/Makefile.partition
PARTITION/Makefile
PARTITION/dtmf_recvr_core/FF
PARTITION/dtmf_recvr_core/FF/vars.tcl
PARTITION/dtmf_recvr_core/FF/EDI
PARTITION/dtmf_recvr_core/FF/EDI/run_all.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_init.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_place.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_prects.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_cts.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_postcts.tcl
October 2014
146

Code Generator
PARTITION/dtmf_recvr_core/FF/EDI/run_postcts_hold.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_route.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_postroute.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_postroute_hold.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_postroute_si_hold.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_postroute_si.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_signoff.tcl
PARTITION/dtmf_recvr_core/FF/EDI/run_debug.tcl
PARTITION/dtmf_recvr_core/FF/procs.tcl
PARTITION/dtmf_recvr_core/Makefile
PARTITION/tdsp_core/FF
PARTITION/tdsp_core/FF/vars.tcl
PARTITION/tdsp_core/FF/EDI
PARTITION/tdsp_core/FF/EDI/run_all.tcl
PARTITION/tdsp_core/FF/EDI/run_init.tcl
PARTITION/tdsp_core/FF/EDI/run_place.tcl
PARTITION/tdsp_core/FF/EDI/run_prects.tcl
PARTITION/tdsp_core/FF/EDI/run_cts.tcl
October 2014
147

Code Generator
PARTITION/tdsp_core/FF/EDI/run_postcts.tcl
PARTITION/tdsp_core/FF/EDI/run_postcts_hold.tcl
PARTITION/tdsp_core/FF/EDI/run_route.tcl
PARTITION/tdsp_core/FF/EDI/run_postroute.tcl
PARTITION/tdsp_core/FF/EDI/run_postroute_hold.tcl
PARTITION/tdsp_core/FF/EDI/run_postroute_si_hold.tcl
PARTITION/tdsp_core/FF/EDI/run_postroute_si.tcl
PARTITION/tdsp_core/FF/EDI/run_signoff.tcl
PARTITION/tdsp_core/FF/EDI/run_debug.tcl
PARTITION/tdsp_core/FF/procs.tcl
PARTITION/tdsp_core/Makefile
The following are the most frequent error and warning messages.
Usage
Example Scripts
Results

October 2014
148

Code Generator

When (required) data is missing or there are syntax errors in the setup.tcl, code generator will abort
with an error message. You can override the error messages by using
set vars(abort) 0
when you set this, an error message will still be issued but the code generator will complete.
A sample error message:
tclsh SCRIPTS/gen_flow.tcl all
-------------------------------------------------
-------------------------------------------------
-----------------------------------------------------
Error Summary
-----------------------------------------------------
(1) A verilog netlist file must be defined
---------------------------------------

When old variables that have changed are being used the system displays this error message so
that the user can update their setup.tcl accordingly.
A sample error message:
Warning Summary
Variable xcap_factor has been changed to post_route_xcap_factor ... please update
October 2014
149

Code Generator
October 2014
150


CPF-Based Low Power Flow is one of the important implementation of the foundation flow.
Before you Begin Implementing CPF-based Low Power Flow
Additional Input
This guide describes the sequence and details of the tasks in the default and recommended
Encounter Digital Implementation System (EDI System) software flow to implement a block or flat
chip using low power techniques such as multiple-supply voltage (MSV) and power shutoff. This flow
assumes that the floorplan is complete (though it does allow the flexibility to implement floorplanning
tasks through plug-in which can be called in the flow). The flow lets you finish the implementation
while,
meeting the timing and physical design requirements,
resolving signal integrity (SI) problems,
and considering on-chip variation (OCV).
The flow also takes advantage of the EDI System software multiple-CPU processing capabilities to
accelerate the design process and, where appropriate, runs features in the multi-threading or
distributed processing mode. As input, you provide a Common Power Format (CPF) file that captures
design and technology-related power constraints. CPF lets you define those rules that EDI System
implements for those objects that the low power design uses. These objects could be: level shifters,
isolation cells, power domains, and power switches. CPF also includes the information EDI System
needs for multi-mode multi-corner (MMMC) timing analysis. With the CPF-Based Low Power
Implementation Flow design, you do not need EDI System commands to define views, corners, and
modes for MMMC.
October 2014
151

For the main design tasks the flow uses the EDI System supercommands with as few nondefault
options as possible.
For a block or a flat chip, these commands are the starting point for the low power design. Ultimately,
each design that you implement might have a different final set of commands, but this flow is the
recommended starting point before customizing a specific design or technology. When you have
completed the tasks described in this guide, you will have a flat MSV design that is ready for DRC
and LVS checks and other sign-off tasks.

The CPF-Based low power implementation flow starts with a completed floorplan (you also have the
flexibility to implement floorplanning tasks through plug-in called in the flow). At the end of each main
task, there is an optimization step, and the task is complete when violations are fixed and the timing
is acceptable. The CPF-Based Low Power flow is complete and the design is ready for sign-off tasks
when the design meets the timing, the CPF-based low power checks are in place, and the physical
design requirements are met.
Before you Begin Implementing CPF-based Low

Power Flow
October 2014
152

The CPF-Based low power implementation flow assumes that timing and routing constraints are
feasible. Make sure the following tasks are complete before starting this flow:
Floorplanning. The information is optional. If you have the floorplan file, you can use it as an
input, or you can perform the floorplanning tasks through the plug-ins in the flow.
Region definition
I/O pad placement
Creation of the clock tree specification file
Creation of RC scale factors
Creation of CPF file
Layer map file for the QRC and GDS export
Config. file for the QRC extraction
For more information, see the following sections of the EDI System User Guide
In the Data Preparation chapter Data Preparation
In the Synthesizing Clock Trees chapter Synthesizing Clock Trees
In the RC Extraction chapter RC Extraction
The following list specifies the input for the CPF-Based low power implementation flow:
EDI System configuration file (setup.tcl) with the following information specified:
SI libraries (*.cdb or *.udn) (recommended but not required)
Clock tree specification file
Scan chain information (*.tcl or *.def)
GDS Layer map file (*.gds)
CPF files (*.cpf)
For more information, see Data Preparation chapter in the EDI System User Guide
October 2014
153

Additional Input
In addition, depending on the design and technology, you may need to supply values for the
following items:
RC scaling factors
Filler cell names
Tie cell names
Welltap cells
Endcap cells
Secondary power cell:pin pair
JTAG instance names
JTAG rows
LSF queue
Dont Use Cells
Delay Cells
CTS Cells
See also,
Defining Variables and Specifying Values for

Command Modes
The CPF-Based Low Power implementation flow uses a set of Tcl scripts that defines the flow
variables and specifies the timing environment, command modes, and metal fill rules. All of the EDI
System Foundation Flows rely on these scripts to keep the flows clean and easy to manage. The
scripts are sourced from the main flow script.
October 2014
154

This file is unique for each low power design. For Low Power design, the user must edit this file
along with the setup.tcl and edi_config.tcl. This file contains Low Power variables such as:
TIE cells for PD
FILLER cells for PD
ENDCAP and WELLTAP implementation
Secondary PG Routing
AON Net synthesis
Modify Power Domains (size, clearance, etc)
Power switch planning
In this section we also cover:
MMMC Timing and SI Setup for Low Power P&R Implementation
See also,
Low power configuration file overlay. This file contains foundation flow variables that are specific to the
LP/CPF flow and should be used in addition to the setup.tcl and edi_config.tcl.
set vars(cpf_keep_rows) true | false
The vars(power_domains) is optional. If not defined, the power domain list will be picked up
automatically.
set vars(power_domains) power domain list
TIE cell information

The variable vars(tie_cells) is defined in the setup.tcl and is used to define a global tie
cell list. This list will be used by default for each power domain. It can be overridden for a given power
domain by setting:
set {{vars(power domain,tie_cells)}} <tie cells for power domain>
October 2014
155

FILLER cell information

The variable vars(filler_cells) is defined in the setup.tcl and is used to define a global filler
cell list. This list will be used by default for each power domain. It can be overridden for a given power
domain by setting:
set vars(power domain,filler_cells) <fillers cells for <power domain>>
Welltap cell information

The variable vars(welltaps) is defined in the setup.tcl and is used to define a global welltap cell
list. This list will be used by default for each power domain. It can be overridden for a given power
domain by setting:
set
set
set
set
set
set
vars(power_domain,welltaps) <welltap cell list for power_domain>

vars(power_domain, welltaps,checkerboard) true or false
vars(power_domain,welltaps,max_gap) max gap in microns
vars(power_domain,welltaps,cell_interval) cell interval in microns
vars(power_domain,welltaps,row_offset) row offset in microns
vars(power_domain, welltaps,verify_rule) verify rule distance
Endcap cell information

The variables vars(pre_endcap) and vars(post_endcap) are set in the setup.tcl file and
are used to define global endcap cells. These used by default for each power domain. They can be
overridden for a given power domain setting:
set vars(power_domain,pre_endcap) <pre endcap cell for power_domain>
set vars(power_domain,post_endcap) <post endcap cell for power_domain>
Always on net buffering

By default SOCE does always-on-net synthesis for SRPG control signal and PSO enable signals as
part of optDesign -preCTS, but this can be manually done for specific nets if necessary. To do
this, define the following variables and uncomment the buffer_always_on_nets proc. In
pre_prects.tcl plug-in template
set
set
set
set
set
set
set
vars(always_on_buffers) list of always on buffers

vars(always_on_nets) list of always on nets
vars(always_on_nets,max_fanout) max fanout limit for always on nets
vars(always_on_nets,max_tran) max transition on always on nets
vars(always_on_nets,max_skew) max skew for always on nets
vars(always_on_nets,max_delay) max delay for always on nets
vars(power_domain,always_on_buffers) <buffers for power domain>
Secondary power/ground routing

Automatic secondary power routing can be enabled during the foundation flows by setting
vars(route_secondary_pg_nets) to true and providing cell pin pair information to identify
the connections requiring routing (Example: AONBUFFD1:TVDD LSL2HCD4:VDDL).
October 2014
156

set vars(route_secondary_pg_nets) [true | false]

set vars(secondary_pg,cell_pin_pairs) secondary power cell pin pair list
In addition, the following can optionally defined either globally or per p/g net
set vars(secondary_pg,max_fanout) max fanout for secondary power routing
set vars(secondary_pg,pattern) secondary power routing pattern trunk | steiner
set vars(secondary_pg,non_default_rule) non-default rule for secondary p/g/
routing
To optionally override for a given p/g net(s), use vars(route_secondary_pg_nets) to

define the list of nets to be overridden and then override vars(<p/g net>,<option>)
set
set
set
set
vars(secondary_pg,nets) list of power/ground nets

vars(p/g_net,max_fanout) max fanout
vars(p/g_net,pattern) trunk | steiner
vars(p/g_net,non_default_rule) non default rule
runCLP options
set vars(clp_options) options for runCLP
Ex: set vars(clp_options) " -cmd ../DATA/chip_top_backend_clp.do extraVlog\../DATA/dummy.v"
Modify power domain

The foundation flow contains set of utilities in utils.tcl, it has modify_power_domains procedure
for implementing the power domain bonding box clearance surrounding to the domain and route
search extensions. For doing power domain modifications user has to define below variables and can
call modify_power_domains procedure through plug-in:
set vars(power_domain,bbox) llx lly urx ury; bondary coordinates in microns
set vars(power_domain,rs_exts) top bot left right; distance in microns
set vars(power_domain,min_gaps) top bot left right; distance in microns
Power Shut-off Planning

The foundation flow contains set of utilities in utils.tcl, it has add_power_switches procedure for
implementing the power switch insertion. User has to define below variables and can call
add_power_switches procedure through plug-in:
October 2014
157

set
set
set
set
set
set
set
set
set
set
set
set
vars(power_domain,switchable) true | false

vars(power_domain,switch_type) column | ring
vars(power_domain,switch_cell) PSO cell name
vars(power_domain,input_enable_pin) PSO cell input enable pin
vars(power_domain,output_enable_pin) PSO cell output enable pin
vars(power_domain,input_enable_net) PSO cell input enable net
vars(power_domain,output_enable_net) PSO cell output enable net
vars(power_domain,switch_instance) switchModuleInstance
vars(power_domain,top_offset) top offset in microns
vars(power_domain,bottom_offset) bottom offset in microns
vars(power_domain,right_offset) right offset in microns
vars(power_domain,left_offset) left offset in microns
Below variables are for column based PSO implementation

set
set
set
set
set
vars(power_domain,checker_board) true | false

vars(power_domain,horizonal_pitch) in microns
vars(power_domain,column_height) Switch cell column height in microns
vars(power_domain,skip_rows) Number of rows to skip
vars(power_domain,back_to_back_chain) true|false
Connects the enableNetOut at the top of a column to the enableNetIn at the top of the next column,
and connects the enableNetOut at the bottom of the column to the enableNetIn at the bottom of the
next column. Below variables are for ring based PSO implementation:
set
set
set
set
vars(power_domain,top_ring) 1|0
vars(power_domain,bottom_ring) 1|0
vars(power_domain,right_ring) 1|0
vars(power_domain,left_ring) 1|0
... defines which side of the power domain to insert switches:

set
set
set
set
vars(power_domain,top_switch_cell) pso cell name

vars(power_domain,bottom_switch_cell) pso cell name
vars(power_domain,left_switch_cell) pso cell name
vars(power_domain,right_switch_cell) pso cell name
... define pso cell name for each side of the power domain
set
set
set
set
set
vars(power_domain,top_filler_cell) filler cell name

vars(power_domain,bottom_filler_cell) filler cell name
vars(power_domain,left_filler_cell) filler cell name
vars(power_domain,right_filler_cell) filler cell name
vars(power_domain,corner_cell_list) corner cell name
... define filler cell name for each side of the power domain
October 2014
158

set
set
set
set
vars(power_domain,top_switches)
vars(power_domain,bottom_switches)
vars(power_domain,left_switches)
vars(power_domain,right_switches)
<"number
<"number
<"number
<"number
-distribute">
-distribute">
-distribute">
-distribute">
... define the number of switches for each side of the power domain
See also,

This is done through the procedure ff_initialize_timing {minmax | mmmc} defined inside
EDI/procs.tcl. Depending on the variation of the flow you are using default, postroute MMMC, or
MMMC the timing environment is initialized to either minmax or mmmc timing. The MMMC timing
environment is required for accurate sign-off timing and is enabled as follows:
In the default flow, it is enabled for the final timing analysis step.
In the postroute MMMC flow, it is enabled after routing.
In the MMMC flow, it is enabled throughout the flow.
See also,
MMMC Timing and SI Setup for Low Power P_R

Implementation
MMMC Timing and SI Setup for Low Power P&R Implementation
Create RC Corners
October 2014
159

create_rc_corner -name $rc_corner

$vars($rc_corner,cap_table) \
-cap_table\
-preRoute_res $vars($rc_corner,pre_route_res_factor)\
-preRoute_cap $vars($rc_corner,pre_route_cap_factor)\
-preRoute_clkres\ $vars($rc_corner,pre_route_clk_res_factor)\
-preRoute_clkcap\ $vars($rc_corner,pre_route_clk_cap_factor)\
-postRoute_res\ $vars($rc_corner,post_route_res_factor)\
-postRoute_cap\ $vars($rc_corner,post_route_cap_factor)\
-postRoute_clkres\ $vars($rc_corner,post_route_clk_res_factor)\
-postRoute_clkcap\ $vars($rc_corner,post_route_clk_cap_factor)\
-postRoute_xcap\ $vars($rc_corner,post_route_xcap_factor)\
-T $vars($rc_corner,T)\
-qx_tech_file $vars($rc_corner, qx_tech_file)\
-qx_conf_file $vars($rc_corner, qx_conf_file) }
October 2014
160

Note: If the low power P&R implementation is not CPF-based or the CPF file does not have its
analysis-views defined, then the user has to create library sets, delay corners, constraint mode and
analysis views.
Please refer to the following section in this guide for these:
See also,
October 2014
161



This section covers the following:
Create Library Sets
See also,
Create Library Sets

foreach library_set $vars(library_sets) {
if {[info exists vars($library_set,si)] && ($vars($library_set,si) != "")} {
create_library_set -name $library_set \
-timing $vars($library_set,timing) \
-si $vars($library_set,si)
} else {
October 2014
162

create_library_set -name $library_set -timing $vars($library_set,timing)
See also,
Create Library Sets

foreach delay_corner $vars(delay_corners) {
set command "create_delay_corner -name $delay_corner "
append command "-library_set $vars($delay_corner,library_set) "
append command "-rc_corner $vars($delay_corner,rc_corner) "
if {[info exists vars($delay_corner,opcond)] && [info exists \

vars($delay_corner,opcond_library)]} {
append command "-opcond_library $vars($delay_corner,opcond_library) "
append command "-opcond $vars($delay_corner,opcond) "
October 2014
163

eval $command
See also,
Create Library Sets

Puts "<FF> Creating constraint modes: $vars(constraint_modes)..."
foreach constraint_mode $vars(constraint_modes) {
create_constraint_mode -name $constraint_mode \
-sdc_files $vars($constraint_mode,pre_cts_sdc)
See also,
Create Library Sets
October 2014
164



set vars(all_analysis_views) [concat $vars(setup_analysis_views) \
$vars(hold_analysis_views)]
Puts "<FF> Creating analysis views: $vars(all_analysis_views) ..."
foreach analysis_view $vars(all_analysis_views) {
if {[lsearch [all_analysis_views] $analysis_view] == -1} {
create_analysis_view -name $analysis_view \
-constraint_mode \ $vars($analysis_view,constraint_mode) \
-delay_corner $vars($analysis_view,delay_corner)
Note: The steps mentioned below are common for CPF and non-CPF based low power P&R
implementation.
See also,
Create Library Sets
October 2014
165


foreach corner $vars(delay_corners) {
Puts "Updating delay corner $corner with RC Corner $vars($corner,rc_corner)\

..."
update_delay_corner -name $corner -rc_corner $vars($corner,rc_corner)
See also,
Create Library Sets

foreach library_set $vars(library_sets) {
if {[info exists vars($library_set,si)] && ($vars($library_set,si) != "")} {
update_library_set -name $library_set \
-si $vars($library_set,si)
}
}
October 2014
166

Set derating factors for delay corners and enable CPPR. CPPR should only be enabled for designs
using derating or designs with specific requirements
set_timing_derate -delay_corner $corner -data -cell_delay -late \

$vars($corner,data_cell_late)
set_timing_derate -delay_corner $corner -data -cell_delay -early \

$vars($corner,data_cell_early)
set_timing_derate -delay_corner $corner -data -net_delay -late \

$vars($corner,data_net_late)
set_timing_derate -delay_corner $corner -data -net_delay -early \

$vars($corner,data_net_early)
set_timing_derate -delay_corner $corner -clock -cell_delay -late \

$vars($corner,clock_cell_late)
set_timing_derate -delay_corner $corner -clock -cell_delay -early \

$vars($corner,clock_cell_early)
set_timing_derate -delay_corner $corner -clock -net_delay -late \

$vars($corner,clock_net_late)
set_timing_derate -delay_corner $corner -clock -net_delay -early \

$vars($corner,clock_net_early)
See also,
Create Library Sets
Define setup and hold Analysis Views and Enable the

October 2014
167

Define setup and hold Analysis Views and Enable the

Default Views
Define setup/hold Analysis Views and Enable the Default Views
set_analysis_view -setup $vars(active_setup_views) -hold $vars(active_hold_views)
set_default_view -setup $vars(default_setup_view) -hold $vars(default_hold_view)
See also,
Create Library Sets

The design is now ready for sign-off physical and timing verification. The output would be:
Design data
Verilog netlist
SPEF file
GDS file
EDI System log file. Use theviewLog
command to open the log file in a new window.
Command file
Reports
Timing reports
Clock tree reports
Verify connectivity
Verify geometry
October 2014
168

Verify metal density

Verify process antenna
Power report
Conformal Low Power report
See also,

Plug-in scripts usage in the reference design:
The following (optional) plug-ins are supported by the flow. They allow for design specific
requirements to be included but allow for flow scripts to remain unchanged. To enable a plug-in,
simply define the appropriate plug-in variable in the setup.tcl, and it will be sourced automatically in
the flow at the appropriate time.
Plugin
Description
always_source_tcl
Sourced at the beginning of every script
pre_partition_tcl
Before partitioning
post_partition
After partitioning
pre_init_tcl
Before design initialization
post_init_tcl
After design initialization
pre_place_tcl
Before placeDesign
place_tcl
Replaces call to placeDesign
post_place_tcl
After placeDesign
pre_cts_tcl
Before clockDesign
cts_tcl
Replaces call to clockDesign
October 2014
169

post_cts_tcl
After clockDesign
pre_route_tcl
Before routeDesign
post_route_tcl
After routeDesign
pre_si_tcl
Before SI fixing
post_si_tcl
After SI fixing
pre_signoff_tcl
Before final STA
post_signoff_tcl
After final STA
The sections below describe some of the plug-in enabled in the reference design.
See also,
CPF File_3a
CPF File:
Define Library Sets and Low Power Cells
set_cpf_version 1.0e
define_library_set -name 0v864 -libraries " \
../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtwc0d720d72.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtwc0d720d72.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpwc0d720d72.lib \
October 2014
170

../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtwc0d72.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtwc0d72.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpwc0d72.lib \
../LIBS/STDCELL/LIBERTY/HVT_CG/tcbn65lphvtcgwc0d72.lib \
../LIBS/STDCELL/LIBERTY/LVT_CG/tcbn65lplvtcgwc0d72.lib \
../LIBS/STDCELL/LIBERTY/NVT_CG/tcbn65lpcgwc0d72.lib \
../LIBS/RAMS_ROMS/LIBERTY/WORST_0864/ulp_dp_512x36cm4sw1bk1.lib \
../LIBS/RAMS_ROMS/LIBERTY/WORST_0864/ulp_rom_512x32cm16.lib \
../LIBS/RAMS_ROMS/LIBERTY/WORST_0864/ulp_sp_2kx32cm8sw1bk2.lib \
../LIBS/IO/LIBERTY/tpzn65lpgv2wc_0864.lib \
../LIBS/ANALOG/LVDS/LIBERTY/lvds_wc_0864.lib \
../LIBS/IP/LIBERTY/pso_wc_0864.lib \
../LIBS/IP/LIBERTY/Module_A_wc_0864.lib \
../LIBS/IP/LIBERTY/Module_B_wc_0864.lib \
../LIBS/IP/LIBERTY/Module_U_wc_0864.lib \
../LIBS/IP/LIBERTY/Module_W_wc_0864.lib \
October 2014
171

define_library_set -name 1v080 -libraries
../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtwc.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtwc.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpwc.lib \
../LIBS/STDCELL/LIBERTY/HVT_CG/tcbn65lphvtcgwc.lib \
../LIBS/STDCELL/LIBERTY/LVT_CG/tcbn65lplvtcgwc.lib \
../LIBS/STDCELL/LIBERTY/NVT_CG/tcbn65lpcgwc.lib \
../LIBS/RAMS_ROMS/LIBERTY/WORST_1080/ulp_dp_512x36cm4sw1bk1.lib \
../LIBS/RAMS_ROMS/LIBERTY/WORST_1080/ulp_rom_512x32cm16.lib \
../LIBS/IO/LIBERTY/tpzn65lpgv2wc.lib \
October 2014
172

../LIBS/ANALOG/PLL/LIBERTY/PL0042.scx3_tsmc_cln80gt_hvt_ss_1p08v_125c.lib \
../LIBS/IP/LIBERTY/pso_wc_1080.lib \
../LIBS/IP/LIBERTY/Module_A_wc_1080.lib \
../LIBS/IP/LIBERTY/Module_B_wc_1080.lib \
../LIBS/IP/LIBERTY/Module_U_wc_1080.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtbc0d881d1.lib \
../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtbc0d881d1.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpbc0d881d1.lib \
../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtbc.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtbc.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpbc.lib \
../LIBS/STDCELL/LIBERTY/HVT_CG/tcbn65lphvtcgbc.lib \
../LIBS/STDCELL/LIBERTY/LVT_CG/tcbn65lplvtcgbc.lib \
../LIBS/STDCELL/LIBERTY/NVT_CG/tcbn65lpcgbc.lib \
October 2014
173

../LIBS/RAMS_ROMS/LIBERTY/BEST_1320/ulp_dp_512x36cm4sw1bk1.lib \
../LIBS/RAMS_ROMS/LIBERTY/BEST_1320/ulp_rom_512x32cm16.lib \
../LIBS/RAMS_ROMS/LIBERTY/BEST_1320/ulp_sp_2kx32cm8sw1bk2.lib \
../LIBS/IO/LIBERTY/tpzn65lpgv2bc.lib \
../LIBS/ANALOG/PLL/LIBERTY/PL0042.scx3_tsmc_cln80gt_hvt_ff_1p32v_0c.lib \
../LIBS/IP/LIBERTY/pso_bc_1320.lib
../LIBS/IP/LIBERTY/Module_A_bc_1320.lib \
../LIBS/IP/LIBERTY/Module_B_bc_1320.lib \
../LIBS/IP/LIBERTY/Module_U_bc_1320.lib \
"
../LIBS/STDCELL/LIBERTY/HVT/tcbn65lphvtbc0d88.lib \
../LIBS/STDCELL/LIBERTY/LVT/tcbn65lplvtbc0d88.lib \
../LIBS/STDCELL/LIBERTY/NVT/tcbn65lpbc0d88.lib \
October 2014
174

../LIBS/STDCELL/LIBERTY/HVT_CG/tcbn65lphvtcgbc0d88.lib \
../LIBS/STDCELL/LIBERTY/LVT_CG/tcbn65lplvtcgbc0d88.lib \
../LIBS/STDCELL/LIBERTY/NVT_CG/tcbn65lpcgbc0d88.lib \
../LIBS/RAMS_ROMS/LIBERTY/BEST_1056/ulp_dp_512x36cm4sw1bk1.lib \
../LIBS/RAMS_ROMS/LIBERTY/BEST_1056/ulp_rom_512x32cm16.lib \
../LIBS/IO/LIBERTY/tpzn65lpgv2bc_1056.lib \
../LIBS/ANALOG/LVDS/LIBERTY/lvds_bc_1056.lib \
../LIBS/IP/LIBERTY/pso_bc_1056.lib
../LIBS/IP/LIBERTY/Module_A_bc_1056.lib \
../LIBS/IP/LIBERTY/Module_B_bc_1056.lib \
../LIBS/IP/LIBERTY/Module_U_bc_1056.lib \
../LIBS/IP/LIBERTY/Module_W_bc_1056.lib \
"
../LIBS/IP/LIBERTY/Module_U.lib \
October 2014
175

"
../LIBS/IP/LIBERTY/Module_U.lib \
"
define_always_on_cell -cells { PTBUFFD8 PTBUFFD4 PTBUFFD2 PTBUFFD1 \
PTBUFFD8HVT PTBUFFD4HVT PTBUFFD2HVT PTBUFFD1HVT \
PTBUFFD8LVT PTBUFFD4LVT PTBUFFD2LVT PTBUFFD1LVT } \
-power_switchable VDD -power TVDD -ground VSS
define_isolation_cell -cells { ISOLOD8 ISOLOD4 ISOLOD2 ISOLOD1 ISOHID8 ISOHID4 \
ISOHID2 ISOHID1 ISOLOD8LVT ISOLOD4LVT ISOLOD2LVT ISOLOD1LVT ISOHID8LVT \
ISOHID4LVT ISOHID2LVT ISOHID1LVT ISOLOD8HVT ISOLOD4HVT ISOLOD2HVT \
ISOLOD1HVT ISOHID8HVT ISOHID4HVT ISOHID2HVT ISOHID1HVT } -power VDD \
-enable
ISO -valid_location to
define_isolation_cell -cells { LVLLHCD8LVT LVLLHCD4LVT LVLLHCD2LVT LVLLHCD1LVT \
LVLLHCD8 LVLLHCD4 LVLLHCD2 LVLLHCD1 LVLLHCD8HVT \
LVLLHCD4HVT LVLLHCD2HVT LVLLHCD1HVT } -power VDD -ground VSS -enable \
NSLEEP -valid_location to
October 2014
176

define_level_shifter_cell -cells { LVLLHCD8LVT LVLLHCD4LVT LVLLHCD2LVT \
LVLLHCD1LVT LVLLHCD8 LVLLHCD4 LVLLHCD2 LVLLHCD1 \
LVLLHCD8HVT LVLLHCD4HVT LVLLHCD2HVT LVLLHCD1HVT } -input_voltage_range \
0.864:1.08:0.216 -output_voltage_range 0.864:1.08:0.216 -direction up \
-input_power_pin VDDL -output_power_pin VDD -ground VSS -enable NSLEEP \
-valid_location to
define_level_shifter_cell -cells { LVLLHD8LVT LVLLHD4LVT LVLLHD2LVT LVLLHD1LVT \
LVLLHD8 LVLLHD4 LVLLHD2 LVLLHD1 LVLLHD8HVT LVLLHD4HVT \
LVLLHD2HVT LVLLHD1HVT } -input_voltage_range 0.864:1.08:0.216 \
-output_voltage_range 0.864:1.08:0.216 -direction up -input_power_pin VDDL \
-output_power_pin VDD -ground VSS -valid_location to
define_power_switch_cell -cells { CDN_RING_SW HDRSID2 HDRSID0 } \
-stage_1_enable NSLEEPIN -stage_1_output NSLEEPOUT -type header \
-power_switchable VDD -power TVDD
define_state_retention_cell -cells { RSDFCSD4 RSDFCSD2 RSDFCSD1 RSDFCRD4 \
RSDFCRD2 RSDFCRD1 RSDFCD4 RSDFCD2 RSDFCD1 RSDFCSD4LVT RSDFCSD2LVT \
RSDFCSD1LVT RSDFCRD4LVT RSDFCRD2LVT RSDFCRD1LVT \
RSDFCD4LVT
October 2014
RSDFCD2LVT RSDFCD1LVT RSDFCSD4HVT \
177

RSDFCSD2HVT RSDFCSD1HVT RSDFCRD4HVT RSDFCRD2HVT \
RSDFCRD1HVT RSDFCD4HVT RSDFCD2HVT RSDFCD1HVT } -clock_pin CP \
-restore_function !NRESTORE -save_function SAVE \
-power_switchable VDD -power TVDD -ground VSS
set_hierarchy_separator /
set_design chip_top
Macro Models
set_macro_model Module_U
create_nominal_condition -name norm_1v080 -voltage 1.08
create_nominal_condition -name ana_2p5 -voltage 2.5
create_nominal_condition -name off -voltage 0
create_power_domain -name PD_3p3v
create_power_domain -name PDdvfs -boundary_ports { Module_U_1* Module_U_2* \
Module_U_3* Module_U_4* Module_U_5* Module_U_6* Module_U_7* Module_U_8* \
Module_U_9* Module_U_10* Module_U_11* Module_U_12* Module_U_13* \
October 2014
178

Module_U_24* Module_U_25* dp dn rrefext Module_U_29* Module_U_30* } \
-default -external_controlled_shutoff
create_power_mode -name PMdvfs_mode1 -default -domain_conditions { \
PD_3p3v@ana_3p3 PD_2p5v@ana_2p5 PDdvfs@norm_1v080 }
create_power_mode -name PMdvfs_mode2 -domain_conditions { PD_3p3v@ana_3p3 \
PD_2p5v@ana_2p5 PDdvfs@norm_0v864 }
create_power_mode -name PM_socoff -domain_conditions { PD_3p3v@ana_3p3 \
PD_2p5v@ana_2p5 PDdvfs@off }
update_power_domain -name PD_3p3v -primary_power_net vdda33 \
-primary_ground_net vssa
update_power_domain -name PDdvfs -primary_power_net VDDdvfs \
-primary_ground_net VSSsoc
end_macro_model
October 2014
179

Create Power Domains

create_power_domain -name PDdvfs -boundary_ports { GPIO_pad[*] SMC_addr_pad[*] \
SMC_data_pad[*] SMC_n_CS_pad[*] SMC_n_be_pad[*] SMC_n_we_pad[*] \
TTC_ext_clk_pad[*] TTC_wfrm_pad[*] spi_N_ss_out_pad[*] } -default \
-external_controlled_shutoff -shutoff_condition !power_on_pin
create_power_domain -name PDpll -instances { i_pll_controller i_PLL \
i_pad_frame/i_pin_12M i_pad_frame/i_pll_clk
i_pad_frame/*PLL_AGND1_H \
i_pad_frame/*PLL_AVDD1_H \
i_pad_frame/*PLL_DGND_i* \
i_pad_frame/*PLL_DVDD_i* \
i_pad_frame/*PLL_AGND0_H \
i_pad_frame/*PLL_AVDD0_H \
*FILLER_PDpll_*} -boundary_ports { \
pin_12M_pad pin_pllclk_pad } -external_controlled_shutoff \
-shutoff_condition !power_on_pin
create_power_domain -name PDw -instances { i_Module_W i_pad_frame/i_lvds \
i_pad_frame/i_rf_rstn i_pad_frame/i_rf_EN i_pad_frame/rf_SW[0].i_rf_SW \
i_pad_frame/rf_SW[1].i_rf_SW i_pad_frame/rf_SW[2].i_rf_SW \
October 2014
180

i_pad_frame/rf_SW[3].i_rf_SW i_pad_frame/*VDDw_i*} -boundary_ports { hissrxip hissrxin \
hissclkp hissclkn hissrxqp hissrxqn rf_resetn_pad rf_en_pad hisstxip \
hisstxin hisstxqp hisstxqn rf_sw_pad[*] } -external_controlled_shutoff \
-shutoff_condition !power_on_pin
create_power_domain -name PDrom -instances { \
i_rom_subsystem/i_rom_wrap/i_rom_core } -shutoff_condition \
!i_apb_subsystem/i_power_ctrl/pwr1_on_rom -secondary_domains { PDdvfs }
create_power_domain -name PDsmc -instances { i_apb_subsystem/i_smc *FILLER_PDsmc_*} \
-shutoff_condition !i_apb_subsystem/i_power_ctrl/pwr1_on_smc \
-secondary_domains { PDdvfs }
create_power_domain -name PDwakeup -instances { i_wakeup \
i_pad_frame/i_shift_en i_pad_frame/i_scan_mode i_pad_frame/i_pin_rst \
i_pad_frame/i_pin_32K i_pad_frame/i_ext_wakeup i_pad_frame/i_wkup_done \
i_pad_frame/i_power_on i_pad_frame/*VDDwakeup_i* *FILLER_PDwakeup_*} -boundary_ports { scan_mode_pad

pin_32k_pad \
pin_reset_pad ext_wakeup_pin wkup_done_pin power_on_pin shift_en_pad }
create_power_domain -name i_apb_subsystem__i_uart1__PDuart_SW -instances { \
October 2014
181

i_apb_subsystem/i_uart1 *FILLER_i_apb_subsystem__i_uart1__PDuart_SW_*} -shutoff_condition \
!i_apb_subsystem/i_power_ctrl/pwr1_on_urt -secondary_domains { PDdvfs }
update_power_domain -name PDdvfs -primary_power_net VDDdvfs \
update_power_domain -name PDpll -primary_power_net VDDpll -primary_ground_net \
VSSsoc
update_power_domain -name PDw -primary_power_net VDDw -primary_ground_net \
VSSsoc
update_power_domain -name PDrom -primary_power_net VDDrom -primary_ground_net \
VSSsoc
update_power_domain -name PDsmc -primary_power_net VDDsmc -primary_ground_net \
VSSsoc
update_power_domain -name PDwakeup -primary_power_net VDDwake \
October 2014
182

update_power_domain -name i_apb_subsystem__i_uart1__PDuart_SW -primary_power_net \
VDDuart -primary_ground_net VSSsoc
Macro Model Mapping

set_instance i_Module_U -model Module_U -domain_mapping { {PDdvfs PDdvfs} \
{PD_3p3v PD_3p3v} {PD_2p5v PD_2p5v} }
Create Power and Ground Nets

create_power_nets -nets VDDdvfs -voltage { 0.864:1.080:0.216 } \
-external_shutoff_condition { !power_on_pin } -peak_ir_drop_limit 0 \
-average_ir_drop_limit 0
create_power_nets -nets VDDpll -voltage 1.08 -external_shutoff_condition { \
!power_on_pin } -peak_ir_drop_limit 0 -average_ir_drop_limit 0
create_power_nets -nets VDDsmc -voltage { 0.864:1.080:0.216 } -internal \
-peak_ir_drop_limit 0 -average_ir_drop_limit 0
create_power_nets -nets VDDrom -voltage { 0.864:1.080:0.216 } -internal \
create_power_nets -nets VDDw -voltage 0.864 -external_shutoff_condition { \
!power_on_pin } -peak_ir_drop_limit 0 -average_ir_drop_limit 0
October 2014
183

create_power_nets -nets VDDwake -voltage 1.08 -peak_ir_drop_limit 0 \
create_power_nets -nets AVDD0_H -voltage 2.5 -peak_ir_drop_limit 0 \
create_power_nets -nets AVDD1_H -voltage 2.5 -peak_ir_drop_limit 0 \
create_power_nets -nets vdda33 -voltage 3.3 -peak_ir_drop_limit 0 \
create_power_nets -nets vdda25 -voltage 2.5 -peak_ir_drop_limit 0 \
create_power_nets -nets VDDuart -voltage { \
0.864:1.080:0.216 } -internal -peak_ir_drop_limit 0 -average_ir_drop_limit \
create_power_nets -nets DUMMY_ESD_VDD -voltage 2.5 -peak_ir_drop_limit 0 \
create_ground_nets -nets AGND0_H -voltage 0.0 -peak_ir_drop_limit 0 \
create_ground_nets -nets AGND1_H -voltage 0.0 -peak_ir_drop_limit 0 \
October 2014
184

create_ground_nets -nets GUARD -voltage 0.0 -peak_ir_drop_limit 0 \
create_ground_nets -nets VSSsoc -voltage 0.00 -peak_ir_drop_limit 0 \
create_ground_nets -nets vssa -voltage 0.00 -peak_ir_drop_limit 0 \
create_ground_nets -nets DUMMY_ESD_VSS -voltage 0.0 -peak_ir_drop_limit 0 \
Create Operating Corners

create_operating_corner -name PDw_wc -voltage 0.864 -process 1 -temperature \
125 -library_set 0v864
create_operating_corner -name PDw_bc -voltage 1.056 -process 1 -temperature 0 \
-library_set 1v056
create_operating_corner -name PDdvfs1_wc -voltage 1.08 -process 1 -temperature \
125 -library_set 1v080
create_operating_corner -name PDdvfs1_bc -voltage 1.320 -process 1 \
-temperature 0 -library_set 1v320
October 2014
185

create_operating_corner -name PDdvfs2_wc -voltage 0.864 -process 1 \
create_operating_corner -name PDdvfs2_bc -voltage 1.056 -process 1 \
create_operating_corner -name PDoff_dvfs1_wc -voltage 0 -process 1 \
create_operating_corner -name PDoff_dvfs1_bc -voltage 0 -process 1 \
create_operating_corner -name PDoff_dvfs2_wc -voltage 0 -process 1 \
create_operating_corner -name PDoff_dvfs2_bc -voltage 0 -process 1 \
Create Nominal Condition

create_nominal_condition -name off -voltage 0
October 2014
186

update_nominal_condition -name norm_1v080 -library_set 1v080
update_nominal_condition -name norm_0v864 -library_set 0v864
update_nominal_condition -name ana_2p5 -library_set 2v5
update_nominal_condition -name ana_3p3 -library_set 3v3
Create Power Modes

create_power_mode -name PMdefault -default -domain_conditions { \
PDpll@norm_1v080 PDwakeup@norm_1v080 PDdvfs@norm_1v080 PDw@norm_0v864 \
PDrom@norm_1v080 PDsmc@norm_1v080 \
i_apb_subsystem__i_uart1__PDuart_SW@norm_1v080 PD_3p3v@ana_3p3 \
PD_2p5v@ana_2p5 }
create_power_mode -name PMdvfs1_smc_off -domain_conditions { PDpll@norm_1v080 \
PDwakeup@norm_1v080 PDdvfs@norm_1v080 PDw@norm_0v864 PDrom@norm_1v080 \
PD_2p5v@ana_2p5 }
create_power_mode -name PMdvfs1_uart_off -domain_conditions { PDpll@norm_1v080 \
PDwakeup@norm_1v080 PDdvfs@norm_1v080 PDw@norm_0v864 PDsmc@norm_1v080 \
PDrom@norm_1v080 i_apb_subsystem__i_uart1__PDuart_SW@off PD_3p3v@ana_3p3 \
PD_2p5v@ana_2p5 }
October 2014
187

create_power_mode -name PMdvfs1_rom_off -domain_conditions { PDpll@norm_1v080 \
PDwakeup@norm_1v080 PDdvfs@norm_1v080 PDw@norm_0v864 PDsmc@norm_1v080 \
PD_2p5v@ana_2p5 }
create_power_mode -name PMdvfs2 -domain_conditions { PDpll@norm_1v080 \
PDwakeup@norm_1v080 PDdvfs@norm_0v864 PDw@norm_0v864 PDrom@norm_0v864 \
PDsmc@norm_0v864 i_apb_subsystem__i_uart1__PDuart_SW@norm_0v864 \
PD_3p3v@ana_3p3 PD_2p5v@ana_2p5 }
create_power_mode -name PMsocoff -domain_conditions { PDwakeup@norm_1v080 \
i_apb_subsystem__i_uart1__PDuart_SW@off PD_3p3v@ana_3p3 PD_2p5v@ana_2p5 }
update_power_mode -name PMdefault -sdc_files \
../DATA/chip_top.sdc
update_power_mode -name PMdvfs2 -sdc_files \
../DATA/chip_top_dvfs2.sdc

create_analysis_view -name AVdefault_BC -mode PMdefault -domain_corners { \
PDdvfs@PDdvfs1_bc PDpll@PDdvfs1_bc PDw@PDw_bc PDrom@PDdvfs1_bc \
PDsmc@PDdvfs1_bc i_apb_subsystem__i_uart1__PDuart_SW@PDdvfs1_bc \
October 2014
188

PDwakeup@PDdvfs1_bc }
create_analysis_view -name AVdefault_WC -mode PMdefault -domain_corners { \
PDdvfs@PDdvfs1_wc PDpll@PDdvfs1_wc PDw@PDw_wc PDrom@PDdvfs1_wc \
PDsmc@PDdvfs1_wc i_apb_subsystem__i_uart1__PDuart_SW@PDdvfs1_wc \
PDwakeup@PDdvfs1_wc }
create_analysis_view -name AVdvfs2_BC -mode PMdvfs2 -domain_corners { \
PDdvfs@PDdvfs2_bc PDpll@PDdvfs1_bc PDw@PDw_bc PDrom@PDdvfs2_bc \
PDsmc@PDdvfs2_bc i_apb_subsystem__i_uart1__PDuart_SW@PDdvfs2_bc \
PDwakeup@PDdvfs1_bc }
create_analysis_view -name AVdvfs2_WC -mode PMdvfs2 -domain_corners { \
PDdvfs@PDdvfs2_wc PDpll@PDdvfs1_wc PDw@PDw_wc PDrom@PDdvfs2_wc \
PDsmc@PDdvfs2_wc i_apb_subsystem__i_uart1__PDuart_SW@PDdvfs2_wc \
PDwakeup@PDdvfs1_wc }
Create Global Connections

create_global_connection -net VDDw -pins { VDD } -domain PDw
create_global_connection -net VSSsoc -pins { VSS } -domain PDw
create_global_connection -net VDDw -pins { VDDw } -domain PDw
create_global_connection -net VSSsoc -pins { VSSsoc } -domain PDw
October 2014
189

create_global_connection -net VDDdvfs -pins { VDD } -domain PDdvfs
create_global_connection -net VSSsoc -pins { VSS } -domain PDdvfs
create_global_connection -net VDDw -pins { VDDL } -domain PDdvfs
create_global_connection -net VDDdvfs -pins { VDD_STATE } -domain PDdvfs
create_global_connection -net VDDdvfs -pins { VDDA } -domain PDdvfs
create_global_connection -net VDDdvfs -pins { TVDD } -domain PDdvfs
create_global_connection -net VDDwake -pins { VDD } -domain PDwakeup
create_global_connection -net VSSsoc -pins { VSS } -domain PDwakeup
create_global_connection -net VDDdvfs -pins { VDDL } -domain PDwakeup
create_global_connection -net VDDpll -pins { VDD } -domain PDpll
create_global_connection -net VSSsoc -pins { VSS } -domain PDpll
create_global_connection -net VDDdvfs -pins { VDDL } -domain PDpll
create_global_connection -net VDDrom -pins { VDD } -domain PDrom
create_global_connection -net VSSsoc -pins { VSS } -domain PDrom
create_global_connection -net VDDdvfs -pins { TVDD } -domain PDrom
create_global_connection -net VDDsmc -pins { VDD } -domain PDsmc
create_global_connection -net VDDdvfs -pins { TVDD } -domain PDsmc
create_global_connection -net VSSsoc -pins { VSS } -domain PDsmc
October 2014
190

Create Low Power Rules

create_power_switch_rule -name pwr_smc_rule -domain PDsmc -external_power_net \
VDDdvfs
create_power_switch_rule -name pwr_rom_rule -domain PDrom -external_power_net \
VDDdvfs
create_power_switch_rule -name i_apb_subsystem__i_uart1__pwr_uart_rule -domain \
i_apb_subsystem__i_uart1__PDuart_SW -external_power_net VDDdvfs
create_isolation_rule -name i_apb_subsystem__i_uart1__iso_uart_hi \
-isolation_condition { i_apb_subsystem/i_power_ctrl/isolate_urt } -pins { \
i_apb_subsystem/i_uart1/ua_nrts } -from { \
i_apb_subsystem__i_uart1__PDuart_SW } -isolation_target from \
-isolation_output high
create_isolation_rule -name i_apb_subsystem__i_uart1__iso_uart_lo \
-isolation_condition { i_apb_subsystem/i_power_ctrl/isolate_urt } -from { \
i_apb_subsystem__i_uart1__PDuart_SW } -exclude { \
i_apb_subsystem/i_uart1/ua_nrts } -isolation_target from -isolation_output \
low
create_isolation_rule -name iso_rom -isolation_condition { \
October 2014
191

i_apb_subsystem/i_power_ctrl/isolate_rom } -from { PDrom } \
-isolation_target from -isolation_output low
create_isolation_rule -name iso_smc_hi -isolation_condition { \
i_apb_subsystem/i_power_ctrl/isolate_smc } -pins { \
i_apb_subsystem/i_smc/smc_n_be[0] i_apb_subsystem/i_smc/smc_n_be[1] \
i_apb_subsystem/i_smc/smc_n_be[2] i_apb_subsystem/i_smc/smc_n_be[3] \
i_apb_subsystem/i_smc/smc_n_cs[0] i_apb_subsystem/i_smc/smc_n_cs[1] \
i_apb_subsystem/i_smc/smc_hready i_apb_subsystem/i_smc/smc_hresp[0] \
i_apb_subsystem/i_smc/smc_n_ext_oe i_apb_subsystem/i_smc/smc_n_rd \
i_apb_subsystem/i_smc/smc_n_we[3] i_apb_subsystem/i_smc/smc_n_we[2] \
i_apb_subsystem/i_smc/smc_n_we[1] i_apb_subsystem/i_smc/smc_n_we[0] \
i_apb_subsystem/i_smc/smc_n_wr } -from { PDsmc } -isolation_target from \
create_isolation_rule -name iso_smc_lo -isolation_condition { \
i_apb_subsystem/i_power_ctrl/isolate_smc } -from { PDsmc } -exclude { \
October 2014
192

i_apb_subsystem/i_smc/smc_hready i_apb_subsystem/i_smc/smc_hresp[0] \
i_apb_subsystem/i_smc/smc_n_be* i_apb_subsystem/i_smc/smc_n_cs* \
i_apb_subsystem/i_smc/smc_n_wr i_apb_subsystem/i_smc/smc_n_we* \
i_apb_subsystem/i_smc/smc_n_rd i_apb_subsystem/i_smc/smc_n_ext_oe } \
-isolation_target from -isolation_output low
create_isolation_rule -name iso_towakeup_hi -isolation_condition { \
i_wakeup/isolate_wkup } -to { PDwakeup } -isolation_target from \
create_level_shifter_rule -name DvfsToPll -from { PDdvfs } -to { PDpll }
create_level_shifter_rule -name DvfsToWakeup -from { PDdvfs } -to { PDwakeup }
create_level_shifter_rule -name Module_WToDvfs -from { PDw } -to { PDdvfs }
create_state_retention_rule -name retn_smc -instances { \
i_apb_subsystem/i_smc/i_apb/i_cfreg0 i_apb_subsystem/i_smc/i_apb/i_cfreg1 \
i_apb_subsystem/i_smc/eco_logic_type4[0].smc_eco_type4 \
i_apb_subsystem/i_smc/eco_logic_type4[1].smc_eco_type4 \
October 2014
193

i_apb_subsystem/i_smc/eco_logic_type4[2].smc_eco_type4 } -restore_edge \
!i_apb_subsystem/i_power_ctrl/nrestore_smc -save_edge \
i_apb_subsystem/i_power_ctrl/save_smc -target_type flop
update_power_switch_rule -name pwr_smc_rule -prefix FE_CPF_PS \
update_power_switch_rule -name pwr_rom_rule -prefix FE_CPF_PS \
update_power_switch_rule -name i_apb_subsystem__i_uart1__pwr_uart_rule -prefix \
FE_CPF_PS -peak_ir_drop_limit 0 -average_ir_drop_limit 0
update_isolation_rules -names i_apb_subsystem__i_uart1__iso_uart_hi -location to \
-prefix CPF_ISO_
update_isolation_rules -names i_apb_subsystem__i_uart1__iso_uart_lo -location to \
-prefix CPF_ISO_
update_isolation_rules -names iso_rom -location to -prefix CPF_ISO_
update_isolation_rules -names iso_smc_hi -location to -prefix CPF_ISO_
update_isolation_rules -names iso_smc_lo -location to -prefix CPF_ISO_
update_isolation_rules -names iso_towakeup_hi -location to -prefix CPF_ISO_
update_level_shifter_rules -names DvfsToPll -location to -prefix CPF_LS_
October 2014
194

update_level_shifter_rules -names DvfsToWakeup -location to -prefix CPF_LS_
update_level_shifter_rules -names Module_WToDvfs -location to -prefix CPF_LS_
post_cts_tcl_3a
post_cts_tcl:
This plug-in is called after CTS inside the run_cts.tcl flow script. You can use this plug-in to adjust IO
timings based on clock insertion delays. The sample commands mentioned below are incorporated in
the post_cts plug-in. You can modify these commands according to your design needs.
set_interactive_constraint_modes [all_constraint_modes -active]
set_propagated_clock [all_clocks]
set_clock_propagation propagated
set_interactive_constraint_modes {}
set_interactive_constraint_modes {PMdefault}
source ../DATA/chip_top_propagated.sdc
set_interactive_constraint_modes {PMdvfs2}
source ../DATA/chip_top_dvfs2_propagated.sdc
post_init_tcl_3a
October 2014
195

post_init_tcl_3a
post_init_tcl:
This plug-in is called after importing the database and CPF inside the run_init.tcl flow script. You can
use this plug-in for floorplan related tasks, which cover:
Die/core boundary creation
Placement of hard macros/blocks
Power domain sizing and clearance surrounding to it
Placement and routing blockages in the floorplan
IO ring creation
And PSO planning
Below are some of the sample commands incorporated in the post_init plug-in. You can modify
these commands according to your design needs:
Define Core Boundary

floorPlan -site core -d 3000 3000 55.0 55.0 55.0 55.0
Relative Placement of Hard Macros/Blocks

relativePlace \
i_sram_subsystem/i_1_sram_voltage_island/i_0_SRAM_2kx32_wrap/i_sram_core \
CORE -orientation R90 -relation L -alignedBy B \
-xSpace 18.0000 -ySpace 20.0000
relativePlace \
October 2014
196

-orientation R270 -relation R -alignedBy B -xSpace 18.0000 -ySpace 0.0000
Power Domain Size and Clearance Information Surrounding it

modifyPowerDomainAttr PDsmc -box
280 1325 440 1550 \
-rsExts 15 15 15 15 -minGaps 15 15 15 15
Column and Ring-based Power Switch Cell Adding

addPowerSwitch -checkerBoard \
-column
-powerDomain i_apb_subsystem__i_uart1__PDuart_SW \
-backToBackChain \
-bottomOffset 60 \
-enableNetIn $n \
-enableNetOut pduart_powered_off \
-enablePinIn NSLEEPIN \
-enablePinOut NSLEEPOUT \
-globalSwitchCellName HDRSID2 \
-leftOffset 20 \
-reportFile $vars(rpt_dir)/PDurt_pso.rpt \
October 2014
197

-switchModuleInstance i_apb_subsystem/i_uart1 \
-topOffset 2 \
-rightOffset 2 \
-horizontalPitch 70 \
-height 125 \
-skipRows 5
addPowerSwitch -powerDomain PDsmc -ring \
-switchModuleInstance i_apb_subsystem/i_smc \
-enableNetIn $n \
-enableNetOut pdsmc_powered_off \
-enablePinIn NSLEEPIN \
-enablePinOut NSLEEPOUT \
-reportFile
$vars(rpt_dir)/PDsmc_pso.rpt \
-cornerCellList CDN_RING_CORNER_UL \
-fillerCellNameBottom CDN_RING_FILLER \
-fillerCellNameLeft CDN_RING_FILLER \
-fillerCellNameRight CDN_RING_FILLER \
-fillerCellNameTop CDN_RING_FILLER \
October 2014
198

-bottomSide 1 \
-leftSide 1 \
-rightSide 1 \
-topSide 1 \
-leftOffset 2 \
-rightOffset 2 \
-topOffset 2 \
-bottomOffset 2 \
-bottomNumSwitch 4 -distribute \
-leftNumSwitch 3 -distribute \
-topNumSwitch 4 -distribute \
-rightNumSwitch 3 -distribute \
-switchCellNameLeft CDN_RING_SW \
-switchCellNameRight CDN_RING_SW \
-switchCellNameTop CDN_RING_SW \
-switchCellNameBottom CDN_RING_SW
Create Halo for Macros

deleteHaloFromBlock -allBlock
October 2014
199

addHaloToBlock 2 2 2 2 i_D0TCM/i_sram_core_1
addHaloToBlock 2 2 2 2 i_ITCM/i_sram_core
Add IO fillers, these are Inserted from Largest to Smallest

addIoFiller -cell PFILLER20
post_place_tcl_3a
post_place_tcl:
This plug-in is called after cell placement inside the run_place.tcl flow script. You can use this plug-in
for standard cell rail creation and check the connectivity and geometry on power nets.
October 2014
200

Standard Cell Rail Creation for Power Domains

Rail Creation for uart Power Domain
sroute -noBlockPins -noPadRings -noPadPins -noStripes \
-straightConnections { straightWithDrcClean straightWithChanges } \
-nets { VDDuart VSSsoc } -powerDomains {i_apb_subsystem__i_uart1__PDuart_SW} \
-targetViaTopLayer 4 \
-crossoverViaTopLayer 4 -verbose
Rail Creation for PDdvfs Power Domain

-nets { VDDdvfs VSSsoc } -powerDomains { PDdvfs }
-targetViaTopLayer 4 -crossoverViaTopLayer 4 -verbose
Rail Creation for PDsmc Power Domain

-nets { VDDsmcVSSsoc } -powerDomains { PDsmc }
-targetViaTopLayer 4 \
-crossoverViaTopLayer 4 -layerChangeTopLayer 4
October 2014
201

Rail Creation for PDwakeup Power Domain

-nets { VDDwake VSSsoc } -powerDomains { PDwakeup }
-targetViaTopLayer 4 -crossoverViaTopLayer 4 -layerChangeTopLayer 4
Check Connectivity and Geometry on Special Nets

verifyConnectivity -type special -nets { VSSsoc VDDdvfs VDDw \
VDDsmc VDDuart VDDwake VDDpll AGND1_H AGND0_H AVDD1_H AVDD0_H \
vdda33 vdda25 vssa}
verifyGeometry -allowPadFillerCellsOverlap \
-allowRoutingBlkgPinOverlap -allowRoutingCellBlkgOverlap
post_prects_tcl_3a
post_prects_tcl:
This plug-in is called after the prects optimization is completed inside the run_prects.tcl flow script.
You can use this plug-in for secondary power pin routing for low power cells (State retention
flops/always-on-buffers/level-shifters) and check the connectivity and geometry on power nets. The
sample commands below are incorporated in the post_prects plug-in. You can modify these
commands according to your design needs.
Note: route_secondary_pg_nets is a procedure implemented in utils.tcl, it does
secondary pg routing.
route_secondary_pg_nets
route_secondary_pg_nets is part of utils.tcl and it internally calls for
October 2014
202

setPGPinUseSignalRoute and routePGPinUseSignalRoute commands for routing

secondary PG nets.
pre_cts_tcl_3a
pre_cts_tcl:
This plug-in is called before CTS inside the run_cts.tcl flow script. You can use this plug-in to define
non-default CTS mode settings. The sample below displays commands incorporated in the pre_cts
plug-in. You can modify these commands according to your design needs.
setCTSMode -topPreferredLayer 9 -bottomPreferredLayer 3
-preferredExtraSpace 2
pre_init_tcl_3a
pre_init_tcl:
This plug-in is called before importing the database, inside the run_init.tcl flow script. You can use this
plug-in for tasks which can be done before importing the design. For example, use this plug-in to insert
buffers on tie-high/tie-low assign statements. The sample commands mentioned below are
incorporated in the pre_init plug-in. You can modify these commands according to your design needs.
setImportMode -bufferTieAssign $vars(buffer_tie_assign)
pre_place_tcl_3a
pre_place_tcl:
This plug-in is called before doing cell placement inside the run_place.tcl flow script.
User can use this plug-in for:
Power planning related tasks which includes:
Power planning for power domains (ring/stripe creations)
Power Shut-off cell power hookup
Welltap/Bias cell power hookup
And commands for enabling always-on-net synthesis as part of preCTS optimization.
Below are some of the sample commands incorporated in the pre_place plug-in. You can modify
these commands according to your design needs.
October 2014
203

Rings Creations for Blocks and for Power Domains

Ring Around PDw Power Domain
selectObject Group PDw
addRing -spacing_bottom 1.0 -width_left 4 -width_bottom 4 \
-width_top 4 -spacing_top 1.0 -layer_bottom M5 \
-stacked_via_top_layer M9 -width_right 4 -around power_domain \
-jog_distance 0.1 -offset_bottom 2 -layer_top M5 -threshold 0.1 \
-offset_left 4 -spacing_right 1.0 -spacing_left 1.0 \
-type block_rings -offset_right 4 -offset_top 4 -layer_right M4 \
-nets { VDDw VSSsoc VDDdvfs} -stacked_via_bottom_layer M1 \
-layer_left M4 -snap_wire_center_to_grid Grid
Ring Around PLL Macro

selectInst i_PLL
-width_top 10 -spacing_top 3.0 -layer_bottom M5 \
-stacked_via_top_layer M9 -width_right 4 -around selected \
-jog_distance 0.1 -offset_bottom 56 -layer_top M5 \
October 2014
204

-threshold 0.1 -offset_left 60 -spacing_right 1.5 \
-spacing_left 1.5 -type block_rings -offset_right 4 \
-offset_top 20 -layer_right M4 -nets {VDDpll VSSsoc} \
-stacked_via_bottom_layer M1 -layer_left M4 \
-snap_wire_center_to_grid Grid
-snap_wire_center_to_grid Grid -width_top 5 -left 0 -top 0 \
-spacing_top 1.5 -layer_bottom M5 -stacked_via_top_layer M5 \
-width_right 5 -around selected -jog_distance 0.1 -offset_bottom 70 \
-layer_top M5 -threshold 0.1 -right 0 -offset_left 78 \
-spacing_right 1.5 -spacing_left 1.5 -type block_rings \
-offset_right 78 -offset_top 78 -layer_right M4 -nets VDDdvfs \
-stacked_via_bottom_layer M1 -layer_left M4
PLL Pad to PLL Ring Connection

sroute -noBlockPins -noPadRings -noCorePins -noStripes \
-padPinAllGeomsConnect -padPinMaxLayer 2 \
-nets {VDDpll} -area { 185 185 2815 2815 }
October 2014
205

PLL Pins to PLL Ring Connection

sroute -blockPinRouteWithPinWidth -noPadRings -noCorePins \
-noPadPins -noStripes -straightConnections { straightWithDrcClean }\
-blockPinTarget { routeToRingOnly } -nets { VDDpll VSSsoc } \
-blocks named -blockNames { i_PLL }
Stripe Creations for Power Domains

PDsmc Power Domain Stripe Creation
selectObject Group PDsmc
addStripe -block_ring_top_layer_limit M5 \
-max_same_layer_jog_length 0.8
-over_power_domain 1 \
-padcore_ring_bottom_layer_limit M1 -set_to_set_distance 70 \
-skip_via_on_pin {} -stacked_via_top_layer M5 \
-padcore_ring_top_layer_limit M5 -spacing 1.6 -xleft_offset 0.905 \
-xright_offset 10 -merge_stripes_value 0.1 -layer M4 \
-block_ring_bottom_layer_limit M1 -width 6.4 -nets \
{VSSsoc VDDdvfs VDDsmc} -stacked_via_bottom_layer M1 \
-snap_wire_center_to_grid Grid
October 2014
206

-max_same_layer_jog_length 0.8 -over_power_domain 1 \
-skip_via_on_pin {} -stacked_via_top_layer M5 \
-padcore_ring_top_layer_limit M5 -spacing 1.6 -merge_stripes_value \
0.1 -direction horizontal -layer M5 -block_ring_bottom_layer_limit M1 \
-width 6.4 -nets {VSSsoc VDDdvfs VDDsmc} -stacked_via_bottom_layer M1 \
-ytop_offset 2 -ybottom_offset 6 -snap_wire_center_to_grid Grid
Full-Chip/Top Level Ring and Stripe Creation

addRing -spacing_bottom 2 -width_left 11 -width_bottom 11
-width_top 11 -spacing_top 2 -layer_bottom M9 \
-stacked_via_top_layer M9 -width_right 11 -around core \
-jog_distance 0.1 -offset_bottom 2.5 -layer_top M9 -threshold 0.1 \
-offset_left 2.5 -spacing_right 2 -spacing_left 2 -offset_right 2.5 \
-offset_top 2.5 -layer_right M8 -nets {VSSsoc VDDdvfs} \
-stacked_via_bottom_layer M8 -layer_leftM8
addRing -spacing_bottom 2 -width_left 11 -width_bottom 11 \
-width_top 11 -spacing_top 2 -layer_bottom M9 \
October 2014
207

-stacked_via_top_layer M9 -width_right 11 -around core \
-jog_distance 0.1 -offset_bottom 28.5 -layer_top M9 \
-threshold 0.1 -offset_left 28.5 -spacing_right 2 -spacing_left 2 \
-offset_right 28.5 -offset_top 28.5 -layer_right M8 -nets \
{VSSsoc VDDdvfs} -stacked_via_bottom_layer M8 -layer_left M8
-max_same_layer_jog_length 0.8 -over_power_domain 1 \
-snap_wire_center_to_grid Grid -padcore_ring_bottom_layer_limit M5 \
-set_to_set_distance 35 -stacked_via_top_layer M9 \
-padcore_ring_top_layer_limit M9 -spacing 1.6 -xleft_offset 54 \
-xright_offset 25 -merge_stripes_value 0.1 -layer M8 \
-block_ring_bottom_layer_limit M5 \
-width 6.4 -nets {VSSsoc VDDdvfs} -stacked_via_bottom_layer M5 \
-allow_jog_block_ring 0
addStripe -area { 193 245 2807 2755 } -block_ring_top_layer_limit M9 \
-max_same_layer_jog_length 0.8 -snap_wire_center_to_grid Grid \
-stacked_via_top_layer M9 -padcore_ring_top_layer_limit M9 \
October 2014
208

-spacing 1.6 -merge_stripes_value 0.1 -direction horizontal -layer M9 \
-block_ring_bottom_layer_limit M5 -width 6.4 -nets {VSSsoc VDDdvfs} \
-stacked_via_bottom_layer M5 -allow_jog_block_ring 0 -ytop_offset 18 \
-ybottom_offset 18
For welltap/Bias/Power-Switch cells PG pin connection, you can run strap over/off-set with regard to
these cells so that the PG connection is established easily. For example, in the below stripe creation
command, the stripe is created over welltap cells (FILLBIAS2A9TH) so that PG connection is easily
established.
addStripe -nets {VDDcpu} -layer ME4 -direction vertical -width 0.2 \
-spacing 1 -over_pins 1 -pin_layer ME1 -master FILLBIAS2A9TH \
-over_power_domain1 -block_ring_top_layer_limit ME7 \
-block_ring_bottom_layer_limit ME4 -stacked_via_top_layer ME7 \
-stacked_via_bottom_layer ME1 -skip_via_on_pin {} \
-switch_layer_over_obs 1
To handle always-on net synthesis as part of preCTS optimization, you have to enable these:
setDontUse PTBUFFD* false
setOptMode -resizeShifterAndIsoInsts true
setvar dpgOptSupportAOFeedThru 1
Add Placement Blockages

createObstruct 2727.626 1098.3065 2755.178 1143.1175
October 2014
209

createObstruct 1992.592 245.018 2012.6265 437.8765
createObstruct 2748.0225 2132.4865 2755.371 2192.543
createObstruct 2750.585 1827.1725 2755.96 1886.794
October 2014
210

Hierarchical Flow
Hierarchical Flow
Hierarchical Flow is a more complex implementation of the flat foundation flow.
Introduction to Hierarchical Implementation Flow

This chapter describes the tasks in the recommended flow using the Encounter Digital
Implementation System (EDI System) software to implement a hierarchical flat chip from a
completed floorplan, while meeting timing and physical design requirements and resolving signal
integrity (SI) problems. In addition, for designs with multiple operating modes or corners that require
optimization, and for more accurate timing, the flow supports multi-mode multi-corner (MMMC)
timing analysis and on-chip variation (OCV) mode. The flow also takes advantage of the EDI
System softwares multiple-CPU processing capabilities to accelerate the design process and runs
features in multi-threading or distributed processing mode where appropriate.
The flow uses the EDI System supercommands. For example, placeDesign, routeDesign,
and optDesign with as few nondefault options as possible for the main design tasks. It is a starting
point for you to use with designs for a hierarchical chip. Ultimately, your design might have a
different final set of commands, but this flow is the recommended starting point, before
customization for your specific design or technology needs.
When the tasks described in this guide are complete, you should have a design that is ready for
DRC and LVS checks and other sign-off tasks.

October 2014
211

Hierarchical Flow

The hierarchical implementation flow starts with a completed floorplan, including partitions. For the
top level and for each block, clock tree estimation is performed for budgeting and CTS.
When implementing a hierarchical flow where the design is divided into partitions, you can treat the partitions
as blackboxes and derive timing estimations for these blackboxes. For a more accurate representation, instead
of using a blackbox at the top level, you can create an Interface Logic Model (ILM) at the block level and use
an ILM as you would use a blackbox.
This flow uses ILMs for the top-level implementation. Steps for using the blackbox flow are noted as
comments but not actually used in the flow.
At the end of each main task, there is an optimization step, and the task is complete when violations are fixed
and the timing is acceptable. The flow is complete and the design is ready for sign-off tasks when the design
meets timing and physical design requirements.
One-Pass and Two -Pass Hierarchical Flows

The hierarchical one-pass ILM flow assumes a single iteration of top-level budgeting. This flow can
be used for designs that do not have complex inter-partition timing flows.
For complex inter-partition timing flows, a single iteration of top-level budgeting might not be
sufficient. In such cases, you can use the hierarchical two-pass ILM flow, which uses a twostep
iteration of top-level budgeting. The two-step flow results in more accurate budgeting and ensures
high accuracy block closure. The two-pass hierarchical flow is provided as reference. The
Foundation Flow kits do not currently provide automation to execute this flow.
Before You Begin - Hierarchical Implementation Flow

The hierarchical implementation flow assumes that timing and routing constraints are feasible.
Make sure the following tasks are complete before starting this flow:
Floorplanning, including the following tasks:
Partition definition
Power grid creation
Region definition
I/O pad placement
October 2014
212

Hierarchical Flow

Note: Leakage power and dynamic power optimization are outside the scope of this flow.
Inputs for Hierarchical Implementation Flow
EDI System configuration file (setup.tcl) with the following information specified:
SI libraries (*.cdb or *.udn) (recommended but not required)
Clock tree specification file
Scan chain information (*.tcl or *.def)In addition, depending on the design and technology,
you may need to supply values for the following items:
RC scaling factors
Filler cell names
Tie cell names
JTAG instance names
JTAG rows
LSF queue
Dont Use Cells
Delay Cells
CTS Cells

October 2014
213

Hierarchical Flow

With the hierarchical partition flow, standard cell placement, feedthrough insertion, pin
assignment, timing budgeting, and partition steps are performed. It provides capabilities for
defining and handling partition blocks, blackboxes, various orientations, rectilinear shapes,
multiply-instantiated partitions, multiple-levels of partitions, partition guard-bands, and so on. It
creates different directories for the top-level and the blocks which can then be independently
implemented by different design teams.
The following capabilities are supported:
Placement or route based partition (feedthrough insertion/pin assignment) flow. You also
have the option to disable feedthrough buffer insertion step.
Channelless and/or design with master/clones.
Option to use prototyping net delay (pico-second per micron) model for quick timing
estimation during partition flow.
FlexModel based hierarchical partition flow for reducing netlist size.
Different virtual optimization engines for timing budgeting
Trial IPO
gigaOptVirtual
Pico-second per micron model (psPM)

October 2014
214

Hierarchical Flow
The hierarchical flow lets you implement the top-level design and the blocks concurrently.
When you partition a design, the EDI System software creates LEF files and budgeted timing
models for the blocks. You can treat the blocks as blackboxes from the top-level perspective
and thus implement the top-level and the blocks concurrently.
During the flow, as the block implementation progresses, the budgeted timing models for the
blocks are often replaced with Interface Logic Models (ILMs) or with more accurate budgeted
timing models. You can also add new LEF files that contain the antenna information for the
blocks.
See also,

An Interface Logic Model (ILM) is a structural representation of the timing characteristics of a block,
specifically a subset of the blocks structure including instances along the I/O timing paths, clocktree instances, and instances or net coupling affecting the signal integrity (SI) on I/O timing paths.
Instead of using a blackbox to represent a block at the top level, you create an ILM at the block level
and use it as you would use a blackbox. Using ILMs to represent the timing characteristics of a
block provides more accurate analysis than using the blackbox flow.
The flow described in this document uses ILMs for representing the partition blocks during the toplevel implementation.
See also,
Hierarchical Flow

October 2014
215

Hierarchical Flow

ILMs can be in either of the following two states:
Flattened state
In this state, the spef file or the Verilog netlist for the ILM is flattened and merged into the top
level. The flattened state is used for accurate delay calculation and Static Timing Analysis.
The flattened state is required by the CTE commands such as report_timing.
Unflattened state
In this state, the spef file or the Verilog netlist of the ILM model is not visible at the top level.
The ILM appears as a blackbox. The unflattened state is used for physical implementation
tasks such as routing.
Here are some commonly used commands that require an unflattened state:
power routing commands such as addStripe
physical cell insertion commands such as addFiller
scan commands such as scanReorder
The EDI System supercommands placeDesign, routeDesign, and
optDesign automatically manage the flattened and unflattened ILM states. For other
commands, you should flatten or unflatten the design as required with the
flattenIlm and the unflattenIlm commands.
See also,
Performing Multi-Mode Multi-Corner Timing Analysis

October 2014
216

Hierarchical Flow
Performing Multi-Mode Multi-Corner Timing Analysis

and Optimization
For designs that contain multiple modes and/or corners, it becomes necessary to configure the
software to support multiple combinations of modes and corners and to evaluate them concurrently.
Multi-Mode Multi-Corner (MMMC) Timing Analysis provides the ability to determine and validate the
timing requirements for such designs.
Multi-Mode Multi-Corner (MMMC) Timing Analysis and Optimization is also recommended for better
timing correlation.
Using MMMC timing analysis, you can concurrently perform the following activities:
Minimum and maximum analysis for each library corner
On-chip variation specification
Push-out and pull-in SDF annotation
Setup analysis uses push-out for launch, pull-in for capture
Hold analysis uses pull-in for launch, and push-out for capture
Note: defineRCCorner flows are not supported with MMMC
Note: For the ILM flow, temperature scaling for resistance must be done using MMMC.
See also,

The ILM models can be used during pre-CTS and post-CTS stages without modification. However,
for blackboxes, the following considerations related to clock latency apply:
Clock latency is reflected directly in the timing arcs. Any included latency is not calculated.
During the pre-CTS stage, the delays of the clock tree are assumed to be 0. To specify clock
latency, create SDC constraints with the set_clock_latency command.
If the delays of the clock tree are included, during pre-CTS optimization, the latency of the
clock pins should be included in the SDC file to account for the top-level clock latency (source
October 2014
217

Hierarchical Flow
latency)
During the post-CTS stage, the delays of the clock tree are calculated and latencies asserted
on pins or with -source are retained.
The deriveTimingBudget command and savePartitioncommand will generate a
budgeted library for use at the top level:
This model assumes zero delay on the clock tree
For pre-CTS optimization, any clock latencies applied to the clock pins of the model
should represent the full clock insertion (top and block)
This model cannot be used in the post-CTS stage. Use the saveModel command to
create new models for each partition in the post-CTS stage.
See also,
Deriving Clock Tree Estimation for Budgeting and

CTS
You need to derive the clock tree estimation for the blocks in the design for timing budgeting and for
Clock Tree Synthesis (CTS). Derive the CTS files and the clock latencies for the top level and for
each block separately as follows:
1. For each block and for the top-level:
create a CTS specification file with the appropriate minimum delay and maximum delay
based on the achievable insertion delays
use the set_clock_latency command to create a latency SDC with for each
clock. The latency should match the minimum- and maximum-delays set in the CTS
specification file.
2. At the top level, specify the latency SDCs with the setPtnUserCnsFile command.
When you perform timing budgeting, latencies in the block- and top-level will be adjusted
based on the latency SDCs. MacroModels for top-level CTS will be used in a .lib flow.
3. Control the latency manually in the following cases:
October 2014
218

Hierarchical Flow
when the design contains clock feedthroughs

when clocks are generated within a partition
4. Derive the budgeted latencies by performing a prototyping pass through the flow including
CTS
Example
The following example illustrates the use flow for deriving CTS files and latencies separately for the
top level and for each block.
For this example, assume a single clock FCLK that attaches to pin pll/clk at the top level and enters
partitions A and B at port clk
The insertion delay target for flat full chip insertion 5 ns. The insertion delay target for partition A is 3
ns and that for partition B is 2 ns.
CTS Specification File

# Ptn A
AutoCTSRootPin clk
MinDelay 3ns
MaxDelay 3ns
# Ptn B
AutoCTSRootPin clk
October 2014
219

Hierarchical Flow
MinDelay 2ns
MaxDelay 2ns
# Top Level
# Only required for Black Box flow
# MacroModel port PtnA/clk 3ns 3ns 3ns 3ns 10ff
# MacroModel port PtnB/clk 2ns 2ns 2ns 2ns 10ff
AutoCTSRootPin pll/clk
MinDelay 5ns
MaxDelay 5ns
Latency SDC File

# Ptn A
set_clock_latency 3 [get_clocks {FCLK}]
# Ptn B
# Full Chip
Note: MacroModels are not required in the ILM flow because CTS automatically calculates the
latency based on the spef file. The saveModel command produces the MacroModels
automatically
October 2014
220

Hierarchical Flow
Latencies for the Full Chip and the Partitions

Before you specify partitions and perform timing budgeting and partitions, the full chip latencies
must either be in the original timing constraints or loaded in a separate SDC
The latency SDCs for each partition are applied using the setPtnUserCnsFile command as
follows
setPtnUserCnsFile \
-fileName PartitionLatencySDCFile
-ptnName PartitionName
In the budgeted SDCs for each partition, the IO arrival and departure times reflect the assigned
latencies. The source and network latencies are also assigned
For this example, the source latency will be 2 ns for partition A (5 ns 3 ns) and 3 ns for partition B
(5 ns -2 ns)
The same budgeted SDCs can be used for pre-CTS and post-CTS optimization. The network
latencies will be ignored once the tree is synthesized
Related Topics:
Creating a Clock Tree Specification File in the Synthesizing Clock Trees chapter in the EDI
System User Guide
Timing Budgeting chapter in the EDI System User Guide
See also,

When working on hierarchical designs, preventing hierarchical signal integrity (SI) issues between
partitions and the top level can become important. You can use many strategies to prevent
hierarchical SI issues. Most of these strategies are performed during floorplanning.
These strategies are not specifically covered in this guide, but include the following:
October 2014
221

Hierarchical Flow
adding placement halos to the partitions to limit the need for routing close to the partition
boundaries
adding routing halos to the partition to limit cross-coupling between routes in the partitions
and top level
Some methodologies such as routing over, or through, partitions using partition feedthroughs
can increase the SI sensitivity. In such cases, you can reduce the SI sensitivity by attaching
buffers to the I/O ports of each partitionthis reduces SI sensitivity by limiting the I/O nets to
short routes and filtering injected noise.
See also,
About the Hierarchical One-Pass ILM Flow and

Diagram
The hierarchical one-pass ILM flow documented in this chapter utilizes a single iteration of top-level
budgeting. This flow is useful for designs that do not have complex inter-partition timing.
For designs that have more complex inter-partition flows, you can use the hierarchical two-pass ILM
flow, which utilizes a two-step iteration of top-level budgeting to generate more accurate budgets
after clock insertion.
The following figure illustrates the hierarchical one-pass ILM flow.
October 2014
222

Hierarchical Flow
See also,
Hierarchical Flow

he hierarchical foundation flow with FlexILM (preCTS) creates reduced netlist and physical model
T
from each partition. With automatic netlist and placement ECO after top-level paths optimization,
memory reduction is done in both timing graph and physical data. With this, the need for strict timing
budget accuracy is reduced, resulting in no requirement for using flattenILM or unflattenILM during
the flow.
October 2014
223

Hierarchical Flow
Setting Up the Flow

To enable the FlexILM (preCTS) hierarchical flow, set vars(enable_flexilm) true
The files and folders listed below are created after tclsh SCRIPTS/gen_flow.tcl m
hier all:
DBS/ FF/ LOG/ make/ Makefile PARTITION/ PARTITION_PRECTS/ RPT/

setup.tcl

October 2014
224

Hierarchical Flow

Make f Makefile
Initial partitioning and budgeting
Run block to preCTS and generate FlexILM
Top FlexILM prectsopt , re-assign ptn pin , timing rebudget and FlexILM-ECO
Assemble updated preCTS DB and check timing
Block implementation from CCOPT/CTS to signoff , generate ILM
Top implementation from CCOPT/CTS to signoff with ILM
Final chip assembly
FlexILM Generation and Top preCTS Optimization with

FlexILM
Block FlexILM Generation After preCTS-opt
createInterfaceLogic -dir block.flexilm -useType flexIlm optStage preCTS
Top preCTS Optimization with FlexILM

###commit flexIlm####
set dirs ""
foreach name $vars(partition_list) { append dirs " -flex_ilm
{${name} ../${name}/${name}.flexilm}" }
eval commit_module_model $dirs -mmmc_file
../../FF/view_definition.tcl
###optDesign ####
setTrialRouteMode -honorPin false -handlePartitionComplex true
setHierMode -optStage preCTS
setOptMode -handlePartitionComplex true
optDesign -preCTS -outDir RPT -prefix flexilm
October 2014
225

Hierarchical Flow
Reassigning PTN Pin and Rebudgeting Timing

###reassign ptn pin and rebudget###
assignPtnPin
checkPinAssignment
setTrialRouteMode -honorPin true
trialRoute
deriveTimingBudget
###flexIlm ECO to update block DB####
set dirs ""
foreach block $vars(partition_list) { append dirs " goldenBlockDir ../${block}/DBS/prects.enc.dat" }
eval update_partition -flexIlmECO $dirs -flexIlmDir
../../PARTITION_FLEXILM -postECOSuffix postECO -pinLocation
See also,
Hierarchical Flow

The design is now ready for sign-off physical and timing verification. The flow outputs the following
files:
Design data
Verilog netlist
SPEF file
GDS file
EDI System log file
Use the viewLog command to open the log file in a new window.
Reports
Timing reports
Clock tree reports
Verify connectivity
October 2014
226

Hierarchical Flow
Verify geometry
Verify metal density
Verify process antenna
See also,
Hierarchical Flow
October 2014
227

Tags for Flow
Tags for Flow

The following list depicts the tags for the EDI System foundation flow.
vars(init,set_distribute_host,tag)
value
vars(init,set_multi_cpu_usage,tag)
value
vars(init,set_rc_factor,tag)
value
vars(init,derate_timing,tag)
value
vars(init,create_rc_corner,tag)
value
vars(init,create_library_set,tag)
value
vars(init,create_delay_corner,tag)
value
vars(init,create_constraint_mode,tag)
value
vars(init,create_analysis_view,tag)
value
vars(init,update_delay_corner,tag)
value
vars(init,update_library_set,tag)
value
vars(init,set_default_view,tag)
value
vars(init,set_power_analysis_mode,tag)
value
October 2014
228

Tags for Flow
vars(init,load_config,tag)
value
vars(init,load_floorplan,tag)
value
vars(init,generate_tracks,tag)
value
vars(init,load_cpf,tag)
value
vars(init,commit_cpf,tag)
value
vars(init,read_activity_file,tag)
value
vars(init,specify_ilm,tag)
value
vars(init,load_ilm_non_sdc_file,tag)
value
vars(init,initialize_timing,tag)
value
vars(init,load_scan,tag)
value
vars(init,specify_spare_gates,tag)
value
vars(init,set_dont_use,tag)
value
vars(init,set_max_route_layer,tag)
value
vars(init,set_design_mode,tag)
value
vars(init,insert_welltaps_endcaps,tag)
value
vars(init,load_config,tag)
value
vars(init,time_design,tag)
value
vars(init,check_design,tag)
value
October 2014
229

Tags for Flow
vars(init,check_timing,tag)
value
vars(init,report_power_domains,tag)
value
vars(place,set_distribute_host,tag)
value
vars(place,set_multi_cpu_usage,tag)
value
vars(place,restore_design,tag)
value
vars(place,initialize_step,tag)
value
vars(place,set_design_mode,tag)
value
vars(place,set_delay_cal_mode,tag)
value
vars(place,set_place_mode,tag)
value
vars(place,set_opt_mode,tag)
value
vars(place,cleanup_specify_clock_tree,tag)
value
vars(place,specify_clock_tree,tag)
value
vars(place,specify_jtag,tag)
value
vars(place,place_jtag,tag)
value
vars(place,place_design,tag)
value
vars(place,add_tie_cells,tag)
value
vars(place,time_design,tag)
value
vars(place,save_design,tag)
value
October 2014
230

Tags for Flow
vars(place,report_power,tag)
value
vars(place,verify_power_domain,tag)
value
vars(place,run_clp,tag)
value
vars(prects,set_distribute_host,tag)
value
vars(prects,set_multi_cpu_usage,tag)
value
vars(prects,initialize_step,tag)
value
vars(prects,set_design_mode,tag)
value
vars(prects,set_ilm_type,tag)
value
vars(prects,cleanup_specify_clock_tree,tag)
value
vars(prects,create_clock_tree_spec,tag)
value
vars(prects,specify_clock_tree,tag)
value
vars(prects,set_useful_skew_mode,tag)
value
vars(prects,set_opt_mode,tag)
value
vars(prects,set_design_mode,tag)
value
vars(prects,set_delay_cal_mode,tag)
value
vars(prects,set_dont_use,tag)
value
vars(prects,opt_design,tag)
value
vars(prects,ck_clone_gate,tag)
value
October 2014
231

Tags for Flow
vars(prects,save_design,tag)
value
vars(prects,report_power,tag)
value
vars(prects,verify_power_domain,tag)
value
vars(prects,run_clp,tag)
value
vars(cts,set_distribute_host,tag)
value
vars(cts,set_multi_cpu_usage,tag)
value
vars(cts,initialize_step,tag)
value
vars(cts,set_design_mode,tag)
value
vars(cts,set_cts_mode,tag)
value
vars(cts,set_nanoroute_mode,tag)
value
vars(cts,enable_clock_gate_cells,tag)
value
vars(cts,clock_design,tag)
value
vars(cts,disable_clock_gate_cells,tag)
value
vars(cts,run_clock_eco,tag)
value
vars(cts,update_timing,tag)
value
vars(cts,time_design,tag)
value
vars(cts,save_design,tag)
value
vars(cts,report_power,tag)
value
October 2014
232

Tags for Flow
vars(cts,verify_power_domain,tag)
value
vars(cts,run_clp,tag)
value
vars(postcts,set_distribute_host,tag)
value
vars(postcts,set_multi_cpu_usage,tag)
value
vars(postcts,initialize_step,tag)
value
vars(postcts,set_design_mode,tag)
value
vars(postcts,set_delay_cal_mode,tag)
value
vars(postcts,set_analysis_mode,tag)
value
vars(postcts,set_opt_mode,tag)
value
vars(postcts,opt_design,tag)
value
vars(postcts,save_design,tag)
value
vars(postcts,report_power,tag)
value
vars(postcts,verify_power_domain,tag)
value
vars(postcts,run_clp,tag)
value
vars(postcts_hold,set_distribute_host,tag)
value
vars(postcts_hold,set_multi_cpu_usage,tag)
value
vars(postcts_hold,initialize_step,tag)
value
vars(postcts_hold,set_dont_use,tag)
value
October 2014
233

Tags for Flow
vars(postcts_hold,set_opt_mode,tag)
value
vars(postcts_hold,opt_design,tag)
value
vars(postcts_hold,save_design,tag)
value
vars(postcts_hold,report_power,tag)
value
vars(postcts_hold,verify_power_domain,tag)
value
vars(postcts_hold,run_clp,tag)
value
vars(route,set_distribute_host,tag)
value
vars(route,set_multi_cpu_usage,tag)
value
vars(route,initialize_step,tag)
value
vars(route,set_nanoroute_mode,tag)
value
vars(route,add_filler_cells,tag)
value
vars(route,route_secondary_pg_nets,tag)
value
vars(route,check_place,tag)
value
vars(route,route_design,tag)
value
vars(route,run_clock_eco,tag)
value
vars(route,spread_wires,tag)
value
vars(route,initialize_timing,tag)
value
vars(route,time_design,tag)
value
October 2014
234

Tags for Flow
vars(route,save_design,tag)
value
vars(route,report_power,tag)
value
vars(route,verify_power_domain,tag)
value
vars(route,run_clp,tag)
value
vars(postroute,set_distribute_host,tag)
value
vars(postroute,set_multi_cpu_usage,tag)
value
vars(postroute,initialize_step,tag)
value
vars(postroute,set_design_mode,tag)
value
vars(postroute,set_extract_rc_mode,tag)
value
vars(postroute,set_analysis_mode,tag)
value
vars(postroute,set_delay_cal_mode,tag)
value
vars(postroute,add_metalfill,tag)
value
vars(postroute,delete_filler_cells,tag)
value
vars(postroute,opt_design,tag)
value
vars(postroute,add_filler_cells,tag)
value
vars(postroute,trim_metalfill,tag)
value
vars(postroute,save_design,tag)
value
vars(postroute,report_power,tag)
value
October 2014
235

Tags for Flow
vars(postroute,verify_power_domain,tag)
value
vars(postroute,run_clp,tag)
value
vars(postroute_hold,set_distribute_host,tag)
value
vars(postroute_hold,set_multi_cpu_usage,tag)
value
vars(postroute_hold,initialize_step,tag)
value
vars(postroute_hold,set_dont_use_mode,tag)
value
vars(postroute_hold,set_opt_mode,tag)
value
vars(postroute_hold,delete_filler_cells,tag)
value
vars(postroute_hold,opt_design,tag)
value
vars(postroute_hold,add_filler_cells,tag)
value
vars(postroute_hold,trim_metalfill,tag)
value
vars(postroute_hold,save_design,tag)
value
vars(postroute_hold,report_power,tag)
value
vars(postroute_hold,verify_power_domain,tag)
value
vars(postroute_hold,run_clp,tag)
value
vars(postroute_si_hold,set_distribute_host,tag)
value
vars(postroute_si_hold,set_multi_cpu_usage,tag)
value
vars(postroute_si_hold,initialize_step,tag)
value
October 2014
236

Tags for Flow
vars(postroute_si_hold,set_design_mode,tag)
value
vars(postroute_si_hold,set_dont_use,tag)
value
vars(postroute_si_hold,set_opt_mode,tag)
value
vars(postroute_si_hold,set_extract_rc_mode,tag)
value
vars(postroute_si_hold,set_si_mode,tag)
value
vars(postroute_si_hold,set_delay_cal_mode,tag)
value
vars(postroute_si_hold,set_analysis_mode,tag)
value
vars(postroute_si_hold,add_metalfill,tag)
value
vars(postroute_si_hold,delete_filler_cells,tag)
value
vars(postroute_si_hold,opt_design,tag)
value
vars(postroute_si_hold,add_filler_cells,tag)
value
vars(postroute_si_hold,trim_metalfill,tag)
value
vars(postroute_si_hold,save_design,tag)
value
vars(postroute_si_hold,report_power,tag)
value
vars(postroute_si_hold,verify_power_domain,tag)
value
vars(postroute_si_hold,run_clp,tag)
value
vars(postroute_si,set_distribute_host,tag)
value
vars(postroute_si,set_multi_cpu_usage,tag)
value
October 2014
237

Tags for Flow
vars(postroute_si,initialize_step,tag)
value
vars(postroute_si,set_design_mode,tag)
value
vars(postroute_si,set_extract_rc_mode,tag)
value
vars(postroute_si,set_si_mode,tag)
value
vars(postroute_si,set_analysis_mode,tag)
value
vars(postroute_si,set_delay_cal_mode,tag)
value
vars(postroute_si,add_metalfill,tag)
value
vars(postroute_si,delete_filler_cells,tag)
value
vars(postroute_si,opt_design,tag)
value
vars(postroute_si,add_filler_cells,tag)
value
vars(postroute_si,trim_metalfill,tag)
value
vars(postroute_si,save_design,tag)
value
vars(postroute_si,report_power,tag)
value
vars(postroute_si,verify_power_domain,tag)
value
vars(postroute_si,run_clp,tag)
value
vars(signoff,set_distribute_host,tag)
value
vars(signoff,set_multi_cpu_usage,tag)
value
vars(signoff,initialize_timing,tag)
October 2014
value
238

Tags for Flow
vars(signoff,initialize_step,tag)
value
vars(signoff,set_analysis_mode,tag)
value
vars(signoff,set_extract_rc_mode,tag)
value
vars(signoff,extract_rc,tag)
value
vars(signoff,dump_spef,tag)
value
vars(signoff,time_design_setup,tag)
value
vars(signoff,time_design_hold,tag)
value
vars(signoff,stream_out,tag)
value
vars(signoff,save_oa_design,tag)
value
vars(signoff,create_ilm,tag)
value
vars(signoff,summary_report,tag)
value
vars(signoff,verify_connectivity,tag)
value
vars(signoff,verify_geometry,tag)
value
vars(signoff,verify_metal_density,tag)
value
vars(signoff,verify_process_antenna,tag)
value
vars(signoff,save_design,tag)
value
vars(signoff,report_power,tag)
value
vars(signoff,verify_power_domain,tag)
value
October 2014
239

Tags for Flow
vars(signoff,ru_clp,tag)
value
vars(partition_list)
value
vars(partition,initialize_timing,tag)
value
vars(partition,load_cpf,tag)
value
vars(partition,commit_cpf,tag)
value
vars(partition,run_clp_init,tag)
value
vars(partition,save_init_dbs,tag)
value
vars(partition,set_budgeting_mode,tag)
value
vars(partition,update_constraint_mode,tag)
value
vars(partition,set_ptn_user_cns_file,tag)
value
vars(partition,set_place_mode,tag)
value
vars(partition,place_design,tag)
value
vars(partition,save_place_dbs,tag)
value
vars(partition,trial_route,tag)
value
vars(partition,assign_ptn_pins,tag)
value
vars(partition,check_pin_assignment,tag)
value
vars(partition,report_unaligned_nets,tag)
value
vars(partition,set_ptn_pin_status,tag)
value
October 2014
240

Tags for Flow
vars(partition,derive_timing_budget,tag)
value
vars(partition,save_budget_dbs,tag)
value
vars(partition,run_clp,tag)
value
vars(partition,partition,tag)
value
vars(partition,save_partition,tag)
value
vars(assemble,assemble_design,tag)
value
vars(assemble,specify_ilm,tag)
value
vars(assemble,load_ilm_non_sdc_file,tag)
value
vars(assemble,load_cpf,tag)
value
vars(assemble,commit_cpf,tag)
value
vars(assemble,initialize_timing,tag)
value
vars(assemble,update_timing,tag)
value
vars(assemble,pre_pac_verify_connectivity,tag)
value
vars(assemble,pre_pac_verify_geometry,tag)
value
vars(assemble,set_module_view,tag)
value
vars(assemble,delete_filler_cells,tag)
value
vars(assemble,opt_design,tag)
value
vars(assemble,add_filler_cells,tag)
value
October 2014
241

Tags for Flow
vars(assemble,post_pac_verify_connectivity,tag)
value
vars(assemble,post_pac_verify_geometry,tag)
value
October 2014
242

10

Executing the Flow

source FF/vars.tcl
source FF/procs.tcl
setDistributeHost -local
setMultiCpuUsage -localCpu max
set vars(step) init
exec mkdir -p $env(VPATH)
See also,
Executing the Flow
Executing the Flow

Executing the Flow
loadConfig FF/run.conf
October 2014
243

generateTracks
readActivityFile -format TCF $vars(activity_file)
create_rc_corner -name rc_min \
-cap_table $vars(rc_min,cap_table) \
-preRoute_res $vars(rc_min,pre_route_res_factor) \
-preRoute_cap $vars(rc_min,pre_route_cap_factor) \
-preRoute_clkres $vars(rc_min,pre_route_clk_res_factor) \
-preRoute_clkcap $vars(rc_min,pre_route_clk_cap_factor) \
-postRoute_res $vars(rc_min,post_route_res_factor) \
-postRoute_cap $vars(rc_min,post_route_cap_factor) \
-postRoute_clkres $vars(rc_min,post_route_clk_res_factor) \
-postRoute_clkcap $vars(rc_min,post_route_clk_cap_factor) \
-postRoute_xcap $vars(rc_min,post_route_xcap_factor) \
-T $vars(rc_min,T) \
-qx_tech_file $vars(rc_min,qx_tech_file)
create_rc_corner -name rc_max \
-cap_table $vars(rc_max,cap_table) \
-preRoute_res $vars(rc_max,pre_route_res_factor) \
October 2014
244

-preRoute_cap $vars(rc_max,pre_route_cap_factor) \
-preRoute_clkres $vars(rc_max,pre_route_clk_res_factor) \
-preRoute_clkcap $vars(rc_max,pre_route_clk_cap_factor) \
-postRoute_res $vars(rc_max,pre_route_res_factor) \
-postRoute_cap $vars(rc_max,pre_route_cap_factor) \
-postRoute_clkres $vars(rc_max,pre_route_clk_res_factor) \
-postRoute_clkcap $vars(rc_max,pre_route_clk_cap_factor) \
-postRoute_xcap $vars(rc_max,pre_route_xcap_factor) \
-T $vars(rc_max,T) \
-qx_tech_file $vars(rc_max,qx_tech_file)
create_library_set \
-name fast \
-timing [list {$vars(fast,timing)} ]
create_library_set \
-name slow \
-timing [list {$vars(slow,timing)} ]
create_delay_corner -name fast_min -library_set fast \
-rc_corner rc_min
October 2014
245

create_delay_corner -name slow_max -library_set slow \
-rc_corner rc_max
create_constraint_mode -name mode1 \
-sdc_files [list {$vars(mode1,pre_cts_sdc)}]
create_constraint_mode -name mode2 \
-sdc_files [list {$vars(mode2,post_cts_sdc)}]
if {[lsearch [all_analysis_views] hold_view] == -1} {
create_analysis_view -name hold_view \
-constraint_mode $vars(hold_view,constraint_mode) \
-delay_corner fast_min
if {[lsearch [all_analysis_views] setup_view] == -1} {
create_analysis_view -name setup_view \
-constraint_mode $vars(setup_view,constraint_mode) \
-delay_corner slow_max
update_library_set -name fast -si [list $vars(fast,si)]
update_library_set -name slow -si [list $vars(slow,si)]
October 2014
246

set active_corners [all_delay_corners]
if {[lsearch $active_corners slow_max] != -1} {
# begin derate.tcl
set_timing_derate -clock \
-cell_delay \
-early \
-delay_corner fast_min $vars(fast_min,clock_cell_early)
set_timing_derate -clock \
-cell_delay \
-late \
-delay_corner slow_max $vars(slow_max,clock_cell_late)
# <end derate.tcl>
set_analysis_view \
-setup [list setup_view] \
-hold [list hold_view]
set_default_view -setup setup_view -hold hold_view
set_power_analysis_mode -analysis_view setup_view
October 2014
247

exec /bin/touch make/init
saveDesign DBS/init.enc -compress -def
report_power -outfile RPT/init.power.rpt
setDontUse $vars(dont_use_list) true
setDontUse $vars(use_list) false
setMaxRouteLayer 6
setDesignMode -process 90
addWellTap -cell [list $vars(welltaps)] \
-prefix WELLTAP \
-maxGap $vars(welltaps,max_gap) \
-inRowOffset $vars(welltaps,row_offset) \
-checkerboard
timeDesign -preplace -prefix preplace -outDir RPT
checkDesign -all check_timing
set vars(step) place
Variables affecting this step are listed here: Place [ on page 49|Example Settings for Each
Script#892148]
setPlaceMode -congEffort $vars(congestion_effort) \
October 2014
248

-clkGateAware $vars(clock_gate_aware) \
-placeIoPins $vars(place_io_pins)
setOptMode -preserveAssertions false \
-leakagePowerEffort $vars(leakage_power_effort) \
-dynamicPowerEffort medium$vars(dynamic_power_effort) \
-criticalRange $vars(critical_range)
setTieHiLoMode -cell {$vars(tie_cells)} \
-maxDistance $vars(tie_cells,max_distance) \
-maxFanout $vars(tie_cells,max_fanout) cleanupSpecifyClockTree\ specifyClockTree -file

$vars(cts_spec) specifyJtag -inst $vars(jtag_cells)\ placeJtag -nrRow $vars(jtag_rows)
placeDesign -inPlaceOpt foreach cell\ {$vars(tie_cells)} {
setDontUse $cell false
addTieHiLo
foreach cell {$vars(tie_cells)} {
setDontUse $cell true
exec /bin/touch make/place
October 2014
249

saveDesign DBS/place.enc -compress
report_power -outfile RPT/place.power.rpt
set vars(step) prects
The variables affecting this step are: PreCTS on page 49
cleanupSpecifyClockTree
specifyClockTree -file $vars(cts_spec)
specifyClockTree -update "AutoCTSRootPin * RouteClkNet YES
setUsefulSkewMode -delayPreCTS true \
-useCells [list $vars(skew_buffers)]
-criticalRange $vars(critical_range) \
-usefulSkew false
exec /bin/touch make/prects
October 2014
250

saveDesign DBS/prects.enc -compress
report_power -outfile RPT/prects.power.rpt
set vars(step) cts
The variables affecting this step are: CTS [ on page 50|Example Settings for Each Script#892076]
setNanoRouteMode \
-routeWithLithoDriven $vars(litho_driven_routing) \
-drouteMultiCutViaEffort $vars(multi_cut_effort)
-criticalRange $vars(critical_range) \
-usefulSkew false
exec /bin/touch make/prects
saveDesign DBS/prects.enc -compress
October 2014
251

report_power -outfile RPT/prects.power.rpt
setNanoRouteMode \
-routeWithLithoDriven $vars(litho_driven_routing) \
-drouteMultiCutViaEffort $vars(multi_cut_effort)
set vars(step) postcts
The variables affecting this step are: PostCTS [ on page 50|Example Settings for Each
Script#892284]
setOptMode -postCtsClkGateCloning true
setAnalysisMode -cppr both
optDesign -postCTS -outDir RPT -prefix postcts
exec /bin/touch make/postcts
saveDesign DBS/postcts.enc -compress
report_power -outfile RPT/postcts.power.rpt
October 2014
252

Flow Setup

Uploaded by

Copyright:

Available Formats

Flow Setup

Uploaded by

Document Information

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

Flow Setup

Uploaded by

Copyright:

Available Formats

EDI System Foundation Flow User Guide

Product Version 14.2

2014-2015 Cadence Design Systems, Inc. All rights reserved.

EDI System Foundation Flow User Guide

About This Manual

Introduction to Foundation Flow

Foundation Flow Overview Diagram

Before You Begin

Input for Running Foundation Flows

Additional Inputs for Running Foundation Flows

About the Flow Documentation

Installing Foundation Flows

Starting Foundation Flows

Defining Floorplanning, Scan, CTS, SDC, and Routing Layer

Defining Hierarchical Partition Information

Defining Library and Technology File

Example of Library and Technology File

Defining RC Corner Information

Example of RC Corner Information

Defining Delay Corner Information

Example of Delay Corner Information

Defining Constraint Modes and Analysis Views

Example of Constraint Modes and Analysis Views

Product Version 14.2

EDI System Foundation Flow User Guide

Mailing Results and Updates

Defining Extraction Options

Defining Noise Settings

Defining Power and Ground Nets

Defining Multi-Threading or Distributed Processing

Example of Distributed Processing

Defining LP.CPF Flow Variables

Defining Optional Variables

Defining Tie and Filler Cells

Defining Welltap and Endcap Specifications

Defining Welltap and Endcap Specifications (Per power domain basis)

Defining Hier ILMs

Defining Flow Control

Defining Command Mode Variables

Defining ccoptDesign for Clock Tree Construction

Defining CCOpt Top and Bottom Layers

Defining CCOpt Variables

Example Settings for Each Script

Plug-in Variables and Tags

Defining Plug-in Variables

Commands that can be Tagged

Tags for EDI System Flow

Product Version 14.2

EDI System Foundation Flow User Guide

Using the Wizard

Starting the Flow Wizard

View and Save Scripts

Setting Up the Library for the Flow Wizard

Adding Library to the Flow - Summary

Setting Up the Library Information

Designing the Flow

Designing the Flow - Netlist

Designing the Flow - Clock Tree

Adding Design to the Flow - Summary

Timing the Flow