Wen-heng Chung
Austin, Texas, United States
Summary
🤩
Rockstar🎓
Top SchoolWen-heng Chung is a veteran software engineer with 12 years of experience building compilers, GPGPU, embedded and server/mobile applications from architecture through delivery. He has led cross-region teams as a solution architect and senior engineering lead, shipping the Clang/LLVM-based HCC compiler used across AMD’s GPUOpen ecosystem and supporting heterogeneous compute stacks. His deep systems and backend expertise shows up in substantial open-source contributions to XLA, ROCm/hip and MIOpen—notably adding AMDGPU support and refactoring GPU compiler and transfer components to work across LLVM backends. Based in Austin, he pairs hands-on kernel and memory-management tuning for ML inference on AMD GPUs with customer-facing roles like pre-sales and field engineering. Colleagues rely on him for stabilizing complex codebases and pragmatic design trade-offs that balance performance, portability and maintainability.
12 years of coding experience
14 years of employment as a software developer
BS, Computer Science, BS, Computer Science at National Chiao Tung University
Chinese, Japanese, English, Spanish
Github Skills (31)
portable10
pytorch10
c-language10
runtimes10
python10
llvm10
run-time10
hiphop10
inference10
hip10
roc10
triton10
compiler-compiler10
deep-learning10
gpu10
Programming languages (9)
Github contributions (5)
ROCm/MIOpen
Sep 2017 - Apr 2022
AMD's Machine Intelligence Library
Role in this project:
Back-end Developer
Back-end DeveloperContributions:42 reviews, 58 commits, 5 PRs in 4 years 7 months
Contributions summary:Wen-heng primarily focused on refactoring and improving the code in the `miopen` repository, which is an AMD's Machine Intelligence Library. Their contributions included reverting a change related to HIP WriteTo/ReadTo, which suggests involvement in the low-level device memory management. The user also addressed code quality by adding `const` qualifiers to APIs, and updating the code related to logging command. These changes indicate a focus on stability and maintainability of the library.
amdcluster-computingmachine-learningblasmachine-intelligence
ROCm/hip
Feb 2016 - May 2019
HIP: C++ Heterogeneous-Compute Interface for Portability
Role in this project:Back-end Developer & Systems Engineer
Back-end Developer & Systems EngineerContributions:46 commits, 30 PRs, 37 pushes in 3 years 3 months
Contributions summary:Wen-heng primarily focused on modifying and maintaining the `hip` library, which is a C++ heterogeneous-compute interface. Contributions include disabling problematic functions, merging branches, adding macro definitions, and fixing compilation errors within the codebase. They also worked on refactoring code and adapting to changes related to promote-free HCC, as well as introducing module implementation and addressing versioning issues, demonstrating a focus on ensuring the library's stability and performance.
cudaheterogeneousgpuportabilityhip-runtime
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.