Haojun Xia
Sydney, New South Wales, Australia
Summary
👤
SeniorHaojun Xia is a Research Software Engineer and PhD candidate in Computer Science at the University of Sydney with nine years’ experience building ML systems and accelerating models on GPUs and FPGAs. His background in computer architecture (USTC) and hands-on research roles at Alibaba Cloud, Together AI, and now Google reflect a blend of low-level systems expertise and large language model research. Haojun has worked across academia and industry—spending time as a visiting researcher in Japan and exchanges in China—which informs a pragmatic, globally minded approach to ML infrastructure. He specializes in turning research prototypes into performant, deployable systems, with particular strength in hardware-aware model optimization. An understated detail: his trajectory shows consistent movement from hardware-focused training to cutting-edge LLM systemization, bridging chip-level understanding with scalable ML engineering.
10 years of coding experience
Github Skills (23)
python10
machine-learning10
deep-learning10
gpu10
numpy9
language-model9
neural-network9
tensor9
gpu-acceleration9
evaluation-framework9
generative-model8
autograd8
computer-architecture7
transformer7
quantization6
Programming languages (5)
Github contributions (5)
Summer-Summer/ComputerArchitectureLab
Mar 2019 - Jul 2019
This repository is used to release the experimental assignments of Computer Architecture Course from USTC
Contributions:22 commits, 6 PRs, 32 pushes in 4 months
computer-architecture
usyd-fsalab/fp6_llm
Mar 2024 - Jul 2025
An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).
Contributions:1 PR, 3 pushes, 1 branch in 1 year 4 months
gpullm-inferencequantization
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.