dse-loop

Automate parameter tuning by running an autonomous design space exploration loop that executes the target program, analyzes results, and iteratively adjusts parameters within a configured time budget. This removes manual trial-and-error when optimizing complex computer architecture or EDA workflows.

Design space exploration handles unknown parameter ranges by automatically inferring them directly from the source code. The workflow then runs baseline and iterative tests to navigate the newly inferred parameter space and optimize the target program effectively.

Yes, autonomous parameter optimization supports configurable TIMEOUT and MAX_ITERATIONS settings. This allows you to define a strict time budget and limit the number of exploration cycles to control compute resource usage during design space tuning.

Recover an interrupted design space exploration workflow using the DSE_STATE.json file. The system features comprehensive recovery-state management, saving the current iteration state and logs to seamlessly resume parameter tuning without losing prior progress.

No, manual range specification is not required for RTL parameter optimization. The exploration workflow can automatically infer parameter ranges from the source code, allowing the iterative tuning process to proceed autonomously without explicit initial values.