Report for NSF Workshop on AI for Electronic Design Automation

The NSF Workshop on AI for Electronic Design Automation (EDA) was held on December 10, 2024, in Vancouver, alongside NeurIPS 2024. It brought together experts across machine learning and EDA to examine how AI—spanning large language models (LLMs), graph neural networks (GNNs), reinforcement learning (RL), neurosymbolic methods, and more—can facilitate EDA and shorten design turnaround. The workshop was organized around four themes: (1) AI for physical synthesis and design for manufacturing (DFM), discussing challenges in physical manufacturing processes and potential AI applications; (2) AI for high-level and logic-level synthesis (HLS/LLS), covering pragma insertion, program transformation, RTL code generation, etc.; (3) AI toolbox for optimization and design, discussing frontier AI developments that could potentially be applied to EDA tasks; and (4) AI for test and verification, including LLM-assisted verification tools, ML-augmented SAT solving, security/reliability challenges, etc. The report recommends NSF to foster AI/EDA collaboration, invest in foundational AI for EDA, develop robust data infrastructures, promote scalable compute infrastructure, and invest in workforce development to democratize hardware design and enable next-generation hardware systems. Workshop information is available at https://ai4eda-workshop.github.io/.

The workshop employed a structured expert-panel format across four thematic tracks, each targeting a critical stage of the EDA flow. Theme 1 addressed physical synthesis and DFM, where AI could optimize layout, placement, routing, and manufacturing yield. Theme 2 focused on HLS/LLS, exploring AI-driven pragma insertion, program transformation, and RTL code generation to raise abstraction and reduce manual effort. Theme 3 surveyed frontier AI developments—such as LLMs, GNNs, RL, and neurosymbolic reasoning—that could be adapted to EDA optimization and design tasks. Theme 4 examined test and verification, including LLM-assisted verification tools, ML-augmented SAT solving, and security/reliability challenges. Discussions were distilled into recommendations for NSF, emphasizing collaboration, foundational AI for EDA, data infrastructures, scalable compute, and workforce development. The workshop website (https://ai4eda-workshop.github.io/) serves as a public repository of materials. The methodology is qualitative and consensus-based, synthesizing expert input rather than reporting new experimental results.