Andrew Kassen
Senior Software Engineer at Intel Corporation
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolAndrew Kassen is a Senior Software Engineer based in San Jose with a PhD in Mathematics from the University of Utah and four years of industry experience focused on high-performance and parallel computing. His doctoral work produced scalable, shared-memory 3D whole-blood simulations on GPUs, combining computational fluid dynamics, solid mechanics, and radial-basis-function elastic cell models—skills he now applies to production-grade performance engineering. At Intel he contributes to GPU-focused back-end work (notably optimizing oneDNN reorder kernels and enabling mixed data types and GPU softmax parity with CPU), bridging numerical methods research and real-world ML library performance. He has a track record of building robust backend systems and parallel algorithms across academia and industry, from immersed-boundary fluid–structure interaction to production web and database stacks. Comfortable in C++, Python and GPU programming, he seeks roles that blend numerical methods, parallel performance tuning, and machine learning at scale. An uncommon strength is his ability to translate deep mathematical models into highly optimized, maintainable code for both research and production environments.
4 years of coding experience
10 years of employment as a software developer
Bachelor of Science - BS, Biochemistry and Molecular Biology, Bachelor of Science - BS, Biochemistry and Molecular Biology at University of California, DavisThe University of Utah
German, English
Github Skills (7)
opencl10
gpu-programming10
oneapi10
c-language10
cprogramming-language10
performance-optimization10
deep-learning8
Programming languages (1)
Github contributions (3)
uxlfoundation/oneDNN
Mar 2022 - Jan 2023
oneAPI Deep Neural Network Library (oneDNN)
Role in this project:
Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:144 reviews, 140 commits, 60 PRs in 10 months
Contributions summary:Andrew primarily focused on improving the performance of the oneDNN library, specifically within the GPU reorder functionality. Their commits addressed issues in the generic reorder kernel, fixing reordering logic, and optimizing memory access patterns. The user also worked on enabling mixed data types and features related to GPU softmax operations, bringing it closer to parity with the CPU implementation. Further contributions included fixes for use as concat, scale handling, and dimension combining.
bfloat16sse41avx512openmpopencl
atkassen/oneDNN
Sep 2024 - Apr 2025
oneAPI Deep Neural Network Library (oneDNN)
Contributions:246 pushes, 46 branches in 6 months
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.