Andrew Kerr
Distinguished Compute Architect at NVIDIA
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolAndrew Kerr is a Distinguished Compute Architect at NVIDIA with 16 years of experience designing high-performance GPU compute systems and software. He progressed through senior technical roles at NVIDIA from 2012 to the present, driving architectures that optimize CUDA kernels and platform build environments. A Georgia Tech-trained PhD in Computer Engineering, he combines deep research rigor with hands-on performance engineering to deliver production-grade compute libraries. His early and foundational contributions to NVIDIA/cutlass—establishing its build system and tuning kernels for CUDA architectures and clang—underscore a practical focus on measurable performance gains. Known for bridging low-level optimization with scalable engineering practices, he operates at the intersection of compiler toolchains, kernel design, and systems integration. Based in San Francisco, he brings rare expertise in both academic methods and industry-scale GPU productization.
16 years of coding experience
13 years of employment as a software developer
Doctor of Philosophy - PhD Computer Engineering, Doctor of Philosophy - PhD Computer Engineering at Georgia Institute of Technology
Github Skills (7)
cuda10
c-language10
cprogramming-language10
gpu10
performance-optimization10
cpp9
deep-learning9
Programming languages (1)
Github contributions (5)
NVIDIA/cutlass
Nov 2017 - Sep 2022
CUDA Templates for Linear Algebra Subroutines
Role in this project:
Back-end Developer / Performance Engineer
Back-end Developer / Performance EngineerContributions:20 releases, 37 reviews, 70 commits in 4 years 10 months
Contributions summary:Andrew's initial commit established the foundation for the project by defining the build environment and CUDA architecture targets. Subsequent commits reflect the user's focus on performance optimization, specifically targeting the CUDA architecture. The user was responsible for integrating and testing different kernels and their performance. Their work included enhancing build configurations to enable compilation with clang, adding various features, and fixing bugs, contributing to the library's overall functionality and performance.
cudacpplinear-algebranvidiamatrix-multiplication
kerrmudgeon/corsairs
Jan 2015 - Jun 2017
Contributions:1 release, 57 pushes, 2 tags in 2 years 5 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.