Chaobo Jia

Biography

Undergrad@HUST
My primary research interest is Systems & Compiler for Machine learning . Nowadays I mainly work on LLM Training & Cloud Systems for ML inference
I am very fortunate to be advised by Minchen Yu(CUHK,SZ), Jiasi Shen(HKUST) and Yao Wan(HUST)

please feel free to chat with me or drop me an email!

B.S. in WuHan, Huazhong University of Science and Technology, Software Engineering 2021.9-2025.6

UCSD Picasso Lab UCSD Sept 2024 - unknown

HKUST Automated Reasoning and Transformation of Software Research Group HongKong July 2024 - Aug. 2024

CUHK,SZ FCS Lab ShenZhen January. 2024 - present

Role: Research Intern, Core author, paper planned for submission to NSDI’25
Advisor: Minchen Yu
Work on Fast Scaling Problem In Serverless GPU Inference Focusing on 0->N.
Introduce Model Parallel Execution Flow to make full use of “free” Resource with incomplete Model Data to increase throughput and reduce cold-start overhead.

University of Leeds & HUST CGCL WuHan Sept. 2023 - Nov. 2023

@CryFin-Tech Company ShangHai, China; Sept 2023 - Dece 2023

Project: tungstenite-rs, rustls
Role: Core Developer
Duties included: Core InferenceEngine Developer, some examples I built:
- Developing a high performance ws protocol library tungstenite-rs-fast
- Refactoring the high-performance openssl protocol library rustls and getting it into production.
- Improving Multi-Node Data Synchronization Performance with RDMA
- Development of a high-performance base library, including high-performance hashTable, mpmc synchronization and other components specialized for specific scenarios.

@ByteDance TikTok BeiJing, China; June 2023 - Octo 2023