AutoPPA framework automates circuit PPA optimization using learned rules

Researchers have developed AutoPPA, a new framework designed to automate the optimization of performance, power, and area (PPA) in RTL design. Unlike previous methods that relied on human-defined rules or lacked prior knowledge, AutoPPA automatically generates its own optimization rules. It achieves this through an Explore-Evaluate-Induce workflow that abstracts rules from contrasting code pairs, leading to more effective optimization patterns. Experiments indicate that AutoPPA surpasses existing state-of-the-art methods and manual optimization techniques. AI

IMPACT Automates RTL design optimization, potentially improving efficiency and reducing manual effort in hardware development.

RANK_REASON This is a research paper introducing a new automated framework for circuit optimization.

Read on arXiv cs.LG →

• AutoPPA
• Chongxiao Li
• LLMs
• RTL
• RTLRewriter
• SymRTLO

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →