About Me
I am a final-year Ph.D. candidate in the Data Science and Analytics Thrust at the Hong Kong University of Science and Technology (Guangzhou), advised by Prof. Xiaowen Chu and co-advised by Prof. Wei Wang (Department of Computer Science and Engineering, HKUST). Prior to my Ph.D. studies, I obtained my M.Sc. and B.Eng. degrees from Peking University and Huazhong University of Science and Technology, respectively.
My research interests lie at the intersection of high-performance computing, computer architecture, and efficient AI systems. I focus on identifying fundamental system bottlenecks and optimization opportunities for emerging AI workloads on modern GPU platforms. Specifically, I design algorithms and systems for:
- Efficient LLM Systems: Optimizing large language model training and inference with hardware-software co-design.
- GNN Systems: Accelerating Graph Neural Networks for large-scale graph learning tasks.
- HPC Applications: Enhancing performance for scientific computing workloads, including graph processing and linear algebra solvers.
My long-term goal is to bridge the gap between theoretical computer architecture and practical, high-performance implementations for real-world AI and scientific computing workloads.
I am currently on the job market looking for Postdoc or Assistant Professor positions.
To Prospective Collaborators
I am always open to collaboration with researchers and students interested in measuring, analyzing, and optimizing large-scale systems. I particularly welcome discussions on efficient GPU kernels, distributed inference frameworks, and system support for sparsity. If you are interested in working with me, please feel free to drop me an email with your background and research interests.
News
11/2025: ZipServ accepted to ACM ASPLOS ’26 (Acceptance rate: 10.6%, 89/840).
11/2025: ROME accepted to ACM PPoPP ’26 (Acceptance rate: 11.5%, 32/280).
10/2025: Started research internship at Alibaba Group (TRE team).
09/2025: SpInfer invited to ACM TOCS (under review).
04/2025: SpInfer honored with the Best Paper Award at ACM EuroSys ’25! (Top 2 of 85 accepted papers with an overall acceptance rate 12.2%)
02/2025: SpInfer received three badges (Available, Functional, Reproducible) at EuroSys ’25.
01/2025: SpInfer accepted to ACM EuroSys ’25.
01/2025: STBLLM accepted to ICLR 2025.
01/2025: Hopper Architecture analysis paper released on arXiv.
12/2024: One paper accepted to IEEE IPDPS ’24.
10/2024: Received the DSA Runner-up Research Prize.
02/2024: DTC-SpMM accepted to ACM ASPLOS ’24.
01/2023: One paper on GNN optimization accepted to IEEE IPDPS ’23.