Xin Tan (谭昕) | Computer Science and Engineering, Chinese University of Hong Kong

About Me

TAN, Xin (谭昕) is currently a PhD Candidate @CUHK CSE, supervised by Prof. Henry Xu. Prior to this, he received his B.E from @NWPU CS in July, 2022.
He was born in Yangzhou, a historic and picturesque city in Jiangsu Province, China.

Research Interests

He is broadly interested in systems design for Machine Learning (MLSys) and has published in top-tier venues across systems, networking, and machine learning, including SIGCOMM, ASPLOS, and ICML. His research mainly focuses on the following topics:

Resource Scheduling: Maximizing resource allocation and utilization to efficiently accommodate workloads in clusters.
ML Inference: Designing scalable and high-performance systems for serving large volumes of inference requests efficiently.
Distributed Training: Developing and optimizing strategies for efficient large-scale model training.

News

[05/2026] Silverstone accepted to ACM SIGCOMM’26.
[06/2025] DSV accepted to ACM ASPLOS’26.
[01/2025] Ayo accepted to ACM ASPLOS’25.
[06/2024] Arlo accepted to ACM ICPP’24.
[08/2022] Join CUHK CSE as a PhD student.

Publications

Conference and Workshops

Dynamic Compute and Network Orchestration for Disaggregated RL

Xin Tan, Yicheng Feng, Yu Zhou, Yimin Jiang, Yibo Zhu, Hong Xu

ACM Special Interest Group on Data Communication (SIGCOMM), 2026.

Arxiv
DSV: Exploiting Dynamic Sparsity to Accelerate Large-Scale Video DiT Training

Xin Tan, Yuetao Chen, Yimin Jiang, Xing Chen, Kun Yan, Nan Duan, Yibo Zhu, Daxin Jiang, Hong Xu

ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2026.

PDF Arxiv
(Artifact Available, Functional, Reproduced)
Towards End-to-End Optimization of LLM-based Applications with Ayo

Xin Tan, Yimin Jiang, Yitao Yang, Hong Xu

ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2025.

PDF Arxiv Code
RFabric: A Reconfigurable Network for the Rhythms of Disaggregated RL

Xin Tan, Yicheng Feng, Yu Zhou, Yimin Jiang, Yibo Zhu, Hong Xu

The 1st Frontier AI Systems Workshop (FAISys), 2025

PDF
Arlo: Serving Transformer-based Language Models with Dynamic Input Lengths

Xin Tan, Jiamin Li, Yitao Yang, Jingzong Li, Hong Xu

ACM International Conference on Parallel Processing (ICPP), 2024.

PDF
FOCUS: DLLMs Know How to Tame Their Compute Bound

Kaihua Liang, Xin Tan, An Zhong, Hong Xu, Marco Canini

Forty-third International Conference on Machine Learning (ICML), 2026

Arxiv
MemShare: Memory Efficient Inference for Large Reasoning Models through KV Cache Reuse

Kaiwen Chen, Xin Tan, Minchen Yu, Jingzong Li, Hong Xu

IEEE/ACM International Symposium on Quality of Service (IWQoS), 2026

Arxiv
Frontier: Simulating the Next Generation of LLM Inference Systems

Yicheng Feng, Xin Tan, Kin Hang Sew, Yimin Jiang, Yibo Zhu, Hong Xu

SOSP Workshop on Practical Adoption Challenges of ML for Systems (PACMI), 2025

Arxiv

Education

	B.E: Northwestern Polytechnical University. Period: 2018.9 - 2022.6 Rank: 1/247
	Ph.D.: The Chinese University of Hong Kong. Period: 2022.8 - 2026.6 (expected)