To Prospective Students

Earning a Ph.D. is a challenging pursuit of learning, exploration, and discovery. I value self-motivated students who are genuinely curious about their research and excited by its potential impact. Below are several intriguing questions. If you can envision yourself enjoying exploring and tackling problems in these areas, our group may be a good fit.

• Architecture: How can we tackle the memory wall through better data orchestration within and across devices by identifying and leveraging the key features of FPGAs, GPUs, CPUs, and emerging hardware?
• Compiler: How can AI models defined in PyTorch be automatically lowered to heterogeneous systems with formally verified correctness and high hardware efficiency?
• HW/SW Co-design: From a hardware perspective, what will future AI models look like (numerical data types, customized structures, etc.) to achieve lower latency and better energy efficiency?

Most of my work is open-source, providing accessible starting points for learning and further exploration. I provide hands-on support and side-by-side guidance while encouraging students to explore and develop their own research ideas.