Chao Liu

Austin, Texas, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

Chao Liu is a Senior Staff Engineer in Menlo Park with nine years of experience specializing in high-performance parallel computing and GPU-accelerated ML infrastructure. He led development of AMD’s Composable Kernel library and drove algorithm work in MIOpen, delivering low-level tensor and kernel optimizations across FP16/BF16 and 3D convolution workloads. At Meta now, he brings a proven track record of squeezing performance from hardware—benchmarked across platforms via contributions to DeepBench—and a background in aerospace and computational engineering that informs his systems-level thinking. Colleagues rely on him to bridge research and production: he translates numerical methods and parallel-mesh experience into production-grade, performance-portable ML primitives.

9 years of coding experience

Github Skills (22)

opencl10

benchmark10

c-language10

operation10

tensorrt10

gpu-programming10

benchmarking10

rnn-model10

n10

amd10

mask-rcnn10

kernel10

hip10

deep-learning10

tensorflow10

Programming languages (6)

C++RoffMLIRAssemblyPythonFortran

Github contributions (5)

ROCm/composable_kernel

Aug 2021 - Dec 2022

[DEPRECATED] Moved to ROCm/rocm-libraries repo. NOTE: develop branch is maintained as a read-only mirror

Role in this project:

Back-end Developer & Performance Engineer

Contributions:953 reviews, 891 commits, 439 PRs in 1 year 3 months

Contributions summary:Chao primarily focused on modifying and optimizing the code for the `Composable Kernel` project, which targets machine learning tensor operators. Their contributions involved merging branches and making code changes within the `composable_kernel` directory, suggesting work related to improving the performance of machine learning kernels. The user's edits included changes to header files and code related to reduction operations and XDL operations, potentially for performance enhancement.

kerneldeep-learningprogramming-modelmachine-learningtensor-learning

ROCm/MIOpen

Mar 2018 - Aug 2021

[DEPRECATED] Moved to ROCm/rocm-libraries repo

Role in this project:

Back-end Developer

Contributions:235 reviews, 406 commits, 44 PRs in 3 years 6 months

Contributions summary:Chao's contributions primarily involve implementing and optimizing low-level tensor operations, and kernel implementations within the AMD's Machine Intelligence Library (MIOpen). The commits showcase the modification of kernel code for tensor copying, scaling, and setting, incorporating performance improvements and supporting various data types, including FP16/BF16. Furthermore, their work includes refactoring and enhancing the implementation for efficient handling of 3D convolution operations.

amdcluster-computingmachine-learningblasmachine-intelligence

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.

Request Free Trial