Chunwei Xia ☕️
Chunwei Xia 春伟 夏

Lecturer (Assistant Professor)

About Me

I am a Lecturer (Assistant Professor) at the School of Computer Science at the University of Leeds. I am also a member of the Distributed Systems and Services Group at Leeds. I am a recipient of the 2023 ACM SIGHPC China Doctoral Dissertation Award.

My research broadly covers computer architecture, compilers, heterogeneous computing, distributed systems, and deep learning. My current research interests include two main areas: (1) building system software using large language models, and (2) creating scalable and cost-effective infrastructures for AI applications:

  • Large Language Model for System Software: I am interested in how to use the Large Language Model (LLM) to construct compilers and decompilers to optimize programs and lift binary codes effectively.

  • Performant and cost-effective infrastructures for AI applications: I work to design and implement Deep Learning (DL) Compilers and frameworks to improve the performance and energy efficiency of emerging AI applications on heterogeneous computing platforms.

Download CV
Interests
  • Cost-Effective Infrastructures for AI
  • Heterogeneous Computing
  • Compiler Optimization
Education
  • PhD Artificial Intelligence

    Institute of Computing Technology, Chinese Academic of Sciences

  • BSc Computer Science and Technology

    Tianjin University

📚 My Research

My research focuses on building system software that is cost-effective and easy to use.

Systems for AI
I am working on developing compilers and frameworks to improve the performance of AI applications. This line of research has successfully produced several papers in top-level computer science conferences, including:
[IPDPS'25], ICCAD'24, ASPLOS'24, ATS'21, TACO'19.

AI for Systems
Since 2023, I have been using AI to build system software in a cost-effective way. Several important papers have been published in:
IPDPS'25, ACL'24, ICSE'24.

Fully funded PhD Studentship in Pattern-aware Parallel Software Development. Apply before 31st January 2025.
Fully funded PhD Studentship in Democratizing Large Deep Learning Models.

We are always looking for self-motivated PhD students.
For candidates with a good research track record, full PhD scholarships are available.
Feel free to email me if you are interested.

We are excited to help industry companies solve their real-world problems. Please reach out to collaborate! 😃

Featured Publications
Recent Publications
(2025). Accelerating Tensor-train Decomposition on Graph Neural Networks. IPDPS'25.
(2025). Leveraging Compilation Statistics for Compiler Phase Ordering. IPDPS'25.
(2024). Optimizing Deep Learning Inference via Global Analysis and Tensor Expression. In ASPLOS ‘24.
Recent & Upcoming Talks
Recent News

✅ Manage your projects

Easily manage your projects - create ideation mind maps, Gantt charts, todo lists, and more!