Hyeonsu Lee
Postdoctoral Researcher at Sungkyunkwan University
Gyeonggi, South Korea
Summary
👤
SeniorHyeonsu Lee is a GPU systems researcher and software engineer with 11 years of experience, currently a postdoctoral researcher at Sungkyunkwan University. He specializes in GPU driver-level optimizations and distributed deep learning infrastructure, having implemented preemptive scheduling in GPU drivers and added NVIDIA MIG instance support to PyTorch DDP for improved HPC resource partitioning. His work blends systems programming and static analysis—he built a tool to automatically detect idempotent properties in GPU kernels and contributed to compiler-assisted thread throttling research to reduce cache contention. Hyeonsu also investigates seamless node scaling for high-availability distributed learning, demonstrating a pragmatic focus on production-ready scalability. Based in Gyeonggi, South Korea, he brings both deep academic insight and hands-on engineering experience that bridge low-level GPU behavior and large-scale ML training systems.
11 years of coding experience
Github Skills (12)
grammar10
abstract-syntax-tree10
grammars10
parser-generator10
antlr10
preemptive8
mali3
gpu3
scheduling2
linux1
linux-kernel1
kernel1
Programming languages (1)
Github contributions (5)
Hyunsu-Lee/preemptive_gpu_scheduling
May 2020 - May 2020
Contributions:13 pushes, 3 branches in 1 day
Hyunsu-Lee/gpu_source
May 2015 - Jul 2015
Contributions:5 pushes, 1 branch in 1 month
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.