AI agents are emerging as a transformative computing paradigm, demonstrating remarkable capabilities in decision-making and problem-solving across diverse applications. Building on these advancements, our research focuses on developing effective systems to tackle complex real-world challenges by automating and streamlining traditionally manual processes.

As one example, our recent work leverages AI agents to automate labor-intensive tasks in hardware design, such as Verilog/HLS coding, assertion generation, and simulator development. By reducing reliance on human effort and domain expertise, we aim to make hardware design more efficient and accessible.

You are welcome to join us as we unlock the potential of AI agents to revolutionize hardware engineering and address broader real-world challenges.

Corresponding Publications:

• [Best Paper Award] MG-Verilog @ LAD'24
• AutoAI2C @ TCAD'24
• GPT4AIGChip @ ICCAD'23
• G-CoS @ ICCAD'21
• AutoNBA @ ICML'21