Shengkai Lin

Shengkai Lin (林圣凯)

Ph.D. Candidate

John Hopcroft Center for Computer Science

jefflin [at] sjtu [dot] edu [dot] cn

Biography

Shengkai is a fourth-year joint Ph.D. candidate at the School of Computer Science at Shanghai Jiao Tong University and the Shanghai Innovation Institute, advised by Prof. Shizhen Zhao at SJTU. He received his bachelor’s degree from SJTU.

During his Ph.D., he was a visiting student at the University of Washington, advised by Prof. Arvind Krishnamurthy, where he collaborated closely with Chenxingyu Zhao, Hongtao Zhang, and Yibo Wu. He also completed research internships at Microsoft Research Asia in 2025 (advised by Ran Shu) and at VMware with the Antrea group in 2022.

Research Projects

He works on networked systems, with a focus on optimizing the host network that bridges computation and physical networks. His recent projects span two directions:

RDMA Networking , advancing GPU networking over RDMA for AI workloads:
- SHIFT [APNet24, arXiv] , an RDMA fault-tolerance layer that extends the ''fallback and bypass'' paradigm across RDMA NICs for training workloads.
eBPF-enhanced Linux Kernel Networking , improving cloud-networking performance and programmability with eBPF:
- ONCache [NSDI25] , a cache-based, low-overhead container overlay network implemented with eBPF, delivering near bare-metal performance.
- VEP [NSDI25] , an annotation-guided eBPF verification toolchain that improves eBPF programmability.

[All Publications]

Research Experiences

University of Washington, Seattle (Jul 2025 - Dec 2025)
Visiting Ph.D. Student, advised by Prof. Arvind Krishnamurthy.
Microsoft Research Asia, Beijing (Sep 2024 - Apr 2025)
Research Intern, advised by Ran Shu.
VMware, Beijing (Jan 2022 - Jun 2022)
Intern, with the Antrea group.