Jake Hemstad
Software Engineering Manager at NVIDIA
Minneapolis, Minnesota, United States
Summary
🤩
Rockstar🎓
Top SchoolJake Hemstad is a software engineering manager at NVIDIA with 10 years of experience building and leading high-performance CUDA C++ libraries, currently heading the CUDA C++ Core Libraries team that maintains Thrust, CUB and libcudacxx. He previously led core development on RAPIDS' libcudf and RMM, contributing low-level memory management and data-structure work that powers GPU DataFrame and allocation tooling used across the ecosystem. Jake pairs hands-on C++14/CUDA optimization and rigorous test automation—his contributions include improving transform_iterator tests and backporting libc++ type-trait tests—with people leadership to make CUDA C++ easier and more reliable for developers. Based in Minneapolis, he blends academic research in scalable/exascale programming models with pragmatic production engineering, a mix that has yielded measurable performance gains and more maintainable libraries.
10 years of coding experience
8 years of employment as a software developer
Bachelor of Arts (B.A.), Computer Science and Physics, Bachelor of Arts (B.A.), Computer Science and Physics at Saint John's UniversityUniversity of Minnesota Twin Cities
Stackoverflow
111reputation
17kreached
2answers
1question
Github Skills (21)
thrust10
algorithms10
libc10
memory-allocation10
c-language10
testing10
memory-management10
gpu-programming10
cudf10
cuda10
cpp10
cprogramming-language10
data-structure9
performance-optimization9
error-handling9
Programming languages (18)
Github contributions (5)
rapidsai/rmm
Dec 2018 - May 2022
RAPIDS Memory Manager
Role in this project:
Back-end Developer
Back-end DeveloperContributions:277 reviews, 616 commits, 89 PRs in 3 years 5 months
Contributions summary:Jake's commits primarily involve the implementation of memory management features within the RAPIDS Memory Manager (RMM) library. Their contributions focus on improving the allocation and deallocation of device memory, including the addition of new memory resource types like CUDA and the implementation of stream-ordered operations. Further contributions included moving and refactoring functions within the codebase. They also added test cases to ensure allocation correctness.
rapidscudamemory-managementmemory-allocation
NVIDIA/cudf
Jun 2018 - Jun 2021
cuDF - GPU DataFrame Library
Role in this project:Back-end Developer
Back-end DeveloperContributions:1866 reviews, 2150 commits, 261 PRs in 3 years
Contributions summary:Jake's contributions centered around the development of the `cudf` library, specifically focusing on enhancing and optimizing its functionality. The commits demonstrate the user's involvement in improving the low-level operations within the library, such as data structure and memory management. Their work involved updating data structures to enable the use of optional data and porting functionality to align with C++14 standard.
cudfdataframegpurapidsarrow
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.