The Design Automation Technical Committee (DATC), is a technical member committee of IEEE Council on Electronic Design Automation (CEDA). The mission of DATC is to serve as a central organization and platform that addresses key challenges in design automation, facilitates collaboration on public design flows and testcases, and organizes relevant workshops, meetings, and publications. In recent years, DATC efforts have focused on building public design flows, defining key metrics, and enabling ML EDA. The overarching goal is to ensure continuous progress in EDA research by maintaining baseline calibrations, tracking metrics, and integrating point CAD tools into a cohesive flow to promote cross-stage research and innovation.
A critical part of DATC activities every year includes maintaining and enhancing a public flow available in this repository called the Robust Design Flow (RDF). IEEE DATC RDF is intended (i) to preserve and integrate leading research codes, including those from past academic contests, and (ii) to provide a foundation and backplane for academic research in the RTL-to-GDS IC implementation arena. Unlike other open-source design flows RDF creates a framework to mix and match point tools at different stages of the flow.
We currently support an academic point tool-based configurable flow that is wrapped around the Single-app integrated OpenROAD flow.
RDF depends on OpenROAD-flow-scripts. Please refer build instructions from ORFS repository and documentation. Once ORFS is built, follow the instructions below to run RDF.
The academic point-tool-based flow is a conventional RDF flow that uses a flow configuration file in YAML format. An example RDF flow configuration file is shown below (Example).
synth:
tool: openroad
user_parms: []
floorplan:
tool: openroad
user_parms: []
global_place:
tool: openroad
user_parms: []
detail_place:
tool: openroad
user_parms: []
cts:
tool: openroad
user_parms: []
global_route:
tool: openroad
user_parms: []
detail_route:
tool: openroad
user_parms: []
finish:
tool: openroad
user_parms: []In the flow configuration, the user can set whether the flow points to the default OpenROAD implementation or a custom point tool.
An example run using the GCD design in OpenROAD for Nangate45 is shown below:
You can run the tool from the command line with the following script.
mkdir run
cd run
python3 ../scripts/robust_design_flow.py -r -c ../scripts/sample_run.yml -d gcd -n nangate45Below we explain the different parameters available for our script.
usage: RobustDesignFlow [-h] [-v | -q] [-t] -c CONFIG [-l LOG] (-r | -p) [-j JOB_ID] -d DESIGN [-m CUSTOM_DESIGN_MAKE] -n PLATFORM
Robust Design Flow (RDF).
options:
-h, --help show this help message and exit
-v, --verbose Enables verbose mode with detailed information.
-q, --quiet Suppresses all informational messages and only displays warnings and errors.
-t, --test Enables test mode override of the defaults
-c CONFIG, --config CONFIG
YAML file containing the target flow
-l LOG, --log_file LOG
Log file for the run.
-r, --run Start a new run
-p, --proceed Continues an existing job or run.
-j JOB_ID, --job_id JOB_ID
Enter the job_id of the run
-d DESIGN, --design DESIGN
Name of the design in ORFS to run
-m CUSTOM_DESIGN_MAKE, --custom_design_make CUSTOM_DESIGN_MAKE
Point to a custom design make file instead of an existing ORFS design
-n PLATFORM, --platform PLATFORM
Define the ORFS platform for the selected design
You can add your own point tool in the RDF configurable flow. First, put your point tool binary and necessary side files in the following directory:
./tools/<stage>/<tool_name>where <stage> is the target design stage, e.g., global placement or detailed routing, and <tool_name> is the name of your tool.
Next, you can add your tool in the flow configuration file, and the RDF Python script will call the custom make command to launch your tool.
An example make command is shown below.
RDF_TRITON_ROUTE_EXE = $(RDF_ROOT_DIR)/../tools/detail_route/TritonRoute/TritonRoute
rdf_run_TritonRoute:
$(RDF_TRITON_ROUTE_EXE) -lef $(TECH_LEF) -def $(RDF_DEF_FILE) -guide $(RESULTS_DIR)/route.guide -output $(RDF_DEF_OUT)Users in the makefile will have access to all the variables that are provided by OpenROAD, such as TECH_LEF, SC_LEF, ADDITIONAL_LEFS, and SDC_FILE. Users must add their tool to the makefile with the prefix rdf_run_.
Additionally, the Python script will provide pointers to the source and destination DEF files or Verilog files for the tool to consume and generate, i.e., RDF_DEF_FILE, RDF_VERILOG_FILE, and RDF_DEF_OUT. Any additional user parameters passed to the YAML file will be passed as Makefile variables with the RDF prefix.
We welcome contributions to DATC RDF. Contributions include:
- Source code of your point tool that can be added as a submodule to our tools directory with the updated Makefile containing the new make target and yaml to support the new tool. If your point tool is distributable, you can also add the binary to our tools directory, along with details in the README on the system configuration in which that binary was built. This should also be accompanied by an update to the Makefile with a new target and the corresponding YAML file.
- New testcases and enablements
To contribute, create a PR to this GitHub repository. Developer certificates will be provided for significant contributions at next year's ICCAD Contest Special Session.
-
V. A. Chhabria, V. Gopalakrishnan, A. B. Kahng, S. Kundu, Z. Wang, B.-Y. Wu and D. Yoon, "Strengthening the Foundations of IC Physical Design and ML EDA Research", Proc. International Conference on Computer-Aided Design, 2024.
-
A. B. Kahng and Z. Wang, "DG-RePlAce: A Dataflow-Driven GPU-Accelerated Analytical Global Placement Framework for Machine Learning Accelerators", IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2024.