Seminar Abstract:

The increasing complexity of integrated circuits (ICs) poses significant efficiency challenges for electronic design automation (EDA), a process that uses computers to implement and verify ICs. Graphics processing units (GPUs), known for their powerful parallel computing capabilities, offer potential solutions to the efficiency challenge by speeding up EDA tools through parallelization. However, designing parallel EDA algorithms that can achieve both high parallelism and high quality is very challenging. This talk will discuss GPU acceleration techniques for various EDA problems, including logic rewriting, global routing and static timing analysis.

About the Speaker:

Shiju Lin received his PhD degree in Computer Science and Engineering from The Chinese University of Hong Kong in 2024. His research interests include GPU acceleration and combinatorial optimization in Electronic Design Automation. He won first place in the CAD Contest at ICCAD 2021 on GPU-Accelerated Logic Rewriting, and won first and second place in the special honor track and main track of the ISPD Contest 2024 on GPU/ML-Enhanced Large-Scale Global Routing.