Sunghyun Park
Sr. System SW Engineer at NVIDIA
United States
Summary
🤩
Rockstar🎓
Top SchoolSunghyun Park is a senior systems software engineer with seven years of experience specializing in compiler optimization, ML systems, auto-tuning, and approximate computing. He holds a PhD in Computer Engineering from the University of Michigan and has driven production-grade ML infrastructure and text-generation services through progressive engineering and management roles at OctoAI and now NVIDIA. An active contributor to the prominent TVM deep learning compiler, he improved MetaSchedule’s measurer, builder/runner, and tuning APIs—work that ties research-grade compiler auto-tuning to real-world deployable toolchains like TensorRT. Comfortable leading small cross-functional teams, he blends deep research experience from academia with hands-on systems engineering to deliver performant ML compilation and serving stacks. Notably, his background includes GPU- and JS-engine performance work from Intel internships, reflecting a long-running focus on low-level performance across platforms.
6 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy - PhD Computer Engineering, Doctor of Philosophy - PhD Computer Engineering at University of MichiganBachelor's degree Electrical and Computer Engineering, Bachelor's degree Electrical and Computer Engineering at Sungkyunkwan University
Github Skills (13)
compiler10
tvm10
machine-learning10
compiler-compiler10
python10
cprogramming-language9
c-language9
tensor9
deep-learning8
performance-analysis8
performance-monitor8
deeplearning-ai8
cuda8
Programming languages (4)
Github contributions (5)
apache/tvm
Dec 2021 - Jan 2023
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:173 reviews, 8 commits, 42 PRs in 1 year 1 month
Contributions summary:Sunghyun contributed to the MetaSchedule component of the TVM project, focusing on improvements to the Measurer module. Their work included modifications to the local builder and runner, as well as updates to the builder input, and integration for TensorRT. Further contributions included the introduction of a tuning API and MetaSchedule passes, and the addition of a shape-to-tensor operation. Their contributions span the build, run, and tuning aspects of the project, demonstrating a focus on compiler and machine learning optimization within the TVM ecosystem.
metalvulkancompilertensoropencl
sunggg/tvm
Nov 2021 - Jun 2024
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Contributions:197 pushes, 52 branches in 2 years 7 months
cpugpu-programminggpu-accelerationtvmdeep-learning
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.