PhD Candidate at Nanjing University
I am a PhD candidate in Computer Science at Nanjing University. My research focuses on computer systems, networking, and AI infrastructure, with an emphasis on practical systems for large-scale production workloads.
State Key Laboratory for Novel Software Technology
Nanjing University · Nanjing, China
I design systems that connect accurate modeling with practical optimization, from production network scheduling to multimodal model inference.
Improves diffusion language model generation efficiency by jointly decoding reliable token groups that have limited dependence on the remaining masked positions.
A cache-aware scheduling system for WeChat's C2C services that improves delivery performance while reducing scheduling variation and traffic cost.
Combines global DNS measurements with HTTP and network ownership evidence to accurately identify websites' CDN providers, including multi-CDN deployments.
Uses reinforcement learning to jointly configure optical links and traffic routes, reducing congestion in hybrid electrical–optical data center networks.
AI Infrastructure Software–Hardware Co-design Intern · Shanghai
Built a high-fidelity simulator for large-model inference that was used for multimodal hardware selection and future system design. Also designed scale-up communication for disaggregated training and inference workloads across vLLM, SGLang, and AReal.
Network Infrastructure Intern · Shenzhen
Developed adaptive scheduling systems for WeChat C2C and real-time communication services. The C2C work led to a publication and patent submission, while RTC optimization reduced call latency by more than 40%.
PhD Candidate, Computer Science and Technology
Bachelor, Computer Science Experimental Program
2023, Second Prize, Future Network Technology Innovation Competition, for work on regional IP anycast deployment and performance.