dse-loop

Automate design space exploration by using a skill that executes programs, parses performance metrics, and iteratively tunes parameters until an objective is met or a timeout occurs. It supports objective-driven optimization and adaptive search strategies for computer architecture configurations.

Tune EDA parameters automatically by running iterative program executions with adaptive search strategies and constraint checking. The process intelligently selects the next parameters to test until the defined optimization objective is met or the specified time budget expires.

Parameter tuning extracts objective metrics through programmatic parsing of program outputs. The skill analyzes these extracted performance metrics, logs the results, and uses them for convergence analysis to guide the next iterative parameter selections.

Yes, you can optimize CPU cache size, associativity, and pipeline width to maximize IPC. The skill runs simulations, checks constraints, and intelligently selects new parameters to test, driving objective-driven optimization for complex computer architecture designs.

No, but the skill requires programmatic parsing of program outputs to extract objective metrics. You must define how the program outputs are parsed so the skill can log results for convergence analysis and intelligently select the next parameters to test.