SUF

Python orchestration for reproducible OpenROAD experiments

SUF orchestrates OpenROAD experiments from Python. A graph of actions and dependencies describes each scenario; asynchronous execution schedules independent work and collects implementation results for comparison.

Overview of the SUF experiment flow.

Flow configuration

SUF uses an external OpenROAD-flow-scripts checkout. Point it at the flow directory:

export SUF_FLOW_ROOT=/path/to/OpenROAD-flow-scripts/flow

OpenROAD and Yosys can be found on PATH or selected with SUF_OPENROAD_EXE and SUF_YOSYS_CMD. Optional HDL translation tools are configured separately. The README describes the environment variables and repository setup.

This separation lets experiments reuse an installed EDA toolchain while Python manages the design variants, run dependencies, and plots. The setup remains a research workflow; reproducing a result requires its scenario and tool versions.

Publication

Louis Ledoux and Marc Casas, The Grafted Superset Approach: Bridging Python to Silicon with Asynchronous Compilation and Beyond, OSDA at DATE, Valencia, 25 March 2024.

HAL record · Poster