POET Optimizes RTL Designs For Lower Power
Researchers introduce POET, a framework using large language models to optimize RTL code for power, performance, and area, submitted Mar 19, 2026. POET uses differential-testing testbench generation to guarantee functional correctness and an LLM-driven evolutionary search with non-dominated sorting and power-first ranking to prioritize power. Evaluated on the RTL-OPT benchmark across 40 designs, it achieves 100% functional correctness and best power on all designs.
Key Points
- 1Achieves 100% functional correctness across 40 RTL designs by using differential testing with golden references
- 2Steers optimization toward low-power Pareto front via LLM-driven evolutionary search and power-first ranking
- 3Enables practitioners to reduce power reliably without manual weight tuning in PPA trade-offs
Scoring Rationale
Strong empirical results and novel LLM-based RTL optimization; limited by single arXiv submission and domain-specific scope.
Sources
Public references used for this report.
Practice with real Logistics & Shipping data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Logistics & Shipping problems
