My research interests lie primarily in operating systems and machine learning systems, with a particular focus on the management of heterogeneous resources:
Heterogeneous Computing: How can we design effective scheduling algorithms and policies for LLM inference on heterogeneous clusters under SLO constraints?
Heterogeneous Memory and Storage: How can operating systems or runtimes adapt to memory access patterns to optimize performance on tiered memory systems (e.g., DRAM + CXL)?
INFO
I'm applying to PhD programs starting in Fall 2026, with a focus on operating systems, memory disaggregation and tiering, CXL, LLM inference, RDMA, and broader systems-related topics. If you're interested in discussing this further, feel free to reach out via email.
📚 Education
- Courses: Operating Systems (Honors Track, A), Computer Organization and Architecture (A), Parallel and Distributed Computing (A), and Compilers (Honors Track, A+).
- Extended program duration by one year to participate in an exchange program at the University of Melbourne.
🏆 Selected Awards
- The highest-level scholarship granted to students in Chinese tertiary education institutions, with a selection rate of ~ 0.4% nationwide.
- Honored as the Outstanding Scholarship Awardee of Fudan University, awarded to one student per department yearly.
📝 Selected Publications
Yi Sun, Zekai Shao, Xingyu Qiu, Yun Li, Ting Liu, Linbing Xiang, Dong Sun, Siming Chen. Interpreting Autonomous Driving Corner Cases: A Visual Analytics Approach. 2024 IEEE 17th Pacific Visualization Conference (PacificVis'24).
Ting Liu, Dong Sun, Chongke Bi, Yi Sun, Siming Chen. Dynamic-Scene-Graph-Supported Visual Understanding of Autonomous Driving Scenarios. 2024 IEEE 17th Pacific Visualization Conference (PacificVis'24).
Lin Gao, Jing Lu, Zekai Shao, Ziyue Lin, Shengbin Yue, Chokit Ieong, Yi Sun, Zhongyu Wei, Siming Chen. Fine‑Tuned Large Language Model for Visualization System: A Study on Self‑Regulated Learning in Education. IEEE Transactions on Visualization and Computer Graphics.
🏫 Teaching
- Migrated the lab platform from Raspberry Pi to the virt generic virtual platform by rewriting the hardware abstraction layer to simplify the environment for students. Developed new lab exercises focused on condition variables and asynchronous I/O.
- Developed labs focusing on Linux features, including mmap, userfaultfd, io_uring and more.