Reproduce “MACO: A Multi-Agent LLM Framework for Automated CGRA Hardware/Software Co-Design” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Zesong Jiang, Yuqi Sun, Qing Zhong, Mahathi Krishna, Deepak Patil, Cheng Tan, Jeff Zhang

MACO automates joint hardware and compiler exploration for coarse-grained reconfigurable arrays by using specialized agents to propose configurations, repair syntax and mapping failures, select candidates, and learn from electronic-design-automation results. Across eight kernels in three domains, power-targeted designs reduce average power by 25.9% versus the best baseline, while speedup falls 7.1%; results depend on optimization target.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.