Chenhao Ye

Chenhao Ye

Research Scientist

ByteDance Seed

I am a research scientist at ByteDance Seed. My research interests span computer systems and AI infrastructure. I currently work on LLM reinforcement learning systems. Please feel free to reach out if you would like to learn more about my research.

Before joining ByteDance Seed, I earned my Ph.D. in computer science from the University of Wisconsin-Madison, where I was advised by Prof. Andrea Arpaci-Dusseau and Prof. Remzi Arpaci-Dusseau as a member of ADSL.

Publications

(2026). TensorHub: Scalable and Elastic Weight Transfer for LLM RL Training. SOSP '26.

PDF Cite URL arXiv

(2026). Cache-Centric Multi-Resource Allocation for Storage Services. FAST '26.

PDF Cite Code Slides URL

(2025). Cloudscape: A Study of Storage Services in Modern Cloud Architectures. FAST '25.

PDF Cite Code URL

(2023). Polaris: Enabling Transaction Priority in Optimistic Concurrency Control. SIGMOD '23.

PDF Cite Code URL

(2023). MadFS: Per-File Virtualization for Userspace Persistent Memory Filesystems. FAST '23.

PDF Cite Code URL

(2022). Using Trātṛ to tame Adversarial Synchronization. USENIX Security '22.

PDF Cite URL

(2021). Scale and Performance in a Filesystem Semi-Microkernel. SOSP '21.

PDF Cite Code URL