IPADS | 并行与分布式系统研究所

张鼎言

张鼎言，2023级直博生，上海交通大学并行与分布式系统研究所。在魏星达老师和陈榕老师的指导下，当前主要研究方向为机器学习系统：聚焦大模型推理服务场景，通过增强算力资源伸缩性，优化请求路由策略，挖掘流量的缓存特征，验证系统架构设计，为实现高性能、低成本的推理提供系统支撑。
I’m Dingyan Zhang, a Ph.D. student at the Institute of Parallel and Distributed Systems (IPADS), Shanghai Jiao Tong University, advised by Xingda Wei and Rong Chen. My current research focuses on machine learning systems, centering on large language model (LLM) serving: I aim to provide system support for high-performance, low-cost inference by enhancing the elasticity of compute resources, optimizing request routing strategies, characterizing cache patterns of request traffic, and validating system architectural specifications.

[OSDI’ 26] Simple is Better: Multiplication May Be All You Need for LLM Request Scheduling. Dingyan Zhang, Jinbo Han, Kaixi Zhang, Xingda Wei, Sijie Shen, Chenguang Fang, Wenyuan Yu, Jingren Zhou, Rong Chen.
[OSDI’ 25] Fast and Live Model Auto Scaling with O(1) Host Caching. Dingyan Zhang, Haotian Wang, Yang Liu, Xingda Wei, Yizhou Shan, Rong Chen, Haibo Chen. [paper]
[FAST’ 26] Fast Cloud Storage for AI Jobs via Grouped I/O API with Transparent Read/Write Optimizations. Yingyi Hao, Ting Yao, Xingda Wei, Dingyan Zhang, Tianle Sun, Yiwen Zhang, Zhiyong Fu, Huatao Wu, Rong Chen.
[ATC’ 25] KVCache Cache in the Wild: Characterizing and Optimizing KVCache Cache at a Large Cloud Provider. Jiahao Wang, Jinbo Han, Xingda Wei, Sijie Shen, Dingyan Zhang, Chenguang Fang, Rong Chen, Wenyuan Yu, Haibo Chen. [paper]