Xinyu Lian
I am a second-year Ph.D. student in Computer Science at the University of Illinois Urbana-Champaign, working with Minjia Zhang. I study elastic and efficient machine learning systems, with a current emphasis on RL training and efficient training kernels.
š Feel free to reach out via email if you are interested in my research.
š„ News
| Jun 25, 2025 | Our work SuperOffload has been accepted at ASPLOS 2026. |
|---|---|
| Apr 29, 2025 | Our work Universal Checkpointing has been accepted at USENIX ATC 2025. |
| Feb 3, 2025 | I am honored to receive the Amazon PhD Fellowship ($70K). |
š Selected Papers
- ASPLOSā26


SuperOffload: Unleashing the Power of Large-Scale LLM Training on SuperchipsProceedings of the 31st ACM International Conference on Architectural Support for Programming Languages and Operating System. 2026Featured by PyTorch Official Blog
Invited for Presentation at PyTorch Conference 2025, Ray x DeepSpeed Meetup
- ATCā25


Universal Checkpointing: A Flexible and Efficient Distributed Checkpointing System for Large-Scale DNN Training with Reconfigurable ParallelismProceedings of the 2025 USENIX Annual Technical Conference. 2025Adopted by Microsoft for Phi-3.5-MoE (42B) Pre-Training
Adopted by BigScience for BLOOM (176B) Pre-Training
Adopted by StepFun and Baidu for multimodal model post-training.
Tested on large-scale clusters (10k+ GPUs) by Oak Ridge and Argonne National Lab
Invited for Presentation at PyTorch Day 2025, FMS 2025, SDC 2025
šļø Service
Organizer: Brett Daniel Software Engineering Seminar
Program Committee: MSR'25
Journal Reviewer: TSE
Artifact Evaluation Committee: SOSP'24