Ching-Hsiang Chu
Menlo Park, California, United States
Summary
🤩
RockstarChing-hsiang Chu is a Senior Software Architect with 11 years of expertise in high-performance and parallel computing, now architecting networking software and systems at NVIDIA in Menlo Park. He has a strong track record building and optimizing MPI stacks and GPU communication layers—work that enabled MPI-based distributed deep learning at supercomputer scale (e.g., ORNL Summit) while keeping one-way latencies low. Previously a research scientist at Meta, he bridged AI and systems co-design to productionize large-scale HPC/AI infrastructure. His Ph.D. work and contributions to MVAPICH2/GDR reflect deep competence in GPU-driven MPI datatype processing and hardware multicast for streaming on massive GPU clusters. Colleagues know him for combining rigorous academic research with pragmatic engineering: he routinely turns million-line codebases into high-performance, production-grade systems. Outside work, his background spans both industry internships and hands-on research roles, including persistent GPU kernel designs for in-memory key-value operations.
11 years of coding experience
Github Skills (41)
pytorch10
benchmark10
python10
evaluation10
acceleration10
mpi10
machine-learning10
benchmarking10
numpy10
deep-learning10
tensorflow10
gpu10
autograd10
neural-network10
cuda10
Programming languages (3)
Github contributions (5)
facebookresearch/torch_ucc
Jul 2021 - Jul 2022
Pytorch process group third-party plugin for UCC
Contributions:45 reviews, 29 commits, 40 PRs in 1 year
pytorchprocess-grouppartythird-party
kingchc/Course-Stuff
Feb 2016 - Aug 2020
Contributions:5 pushes, 1 branch in 4 years 6 months
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.