Rasmus Larsen
Distinguished Engineer at NVIDIA
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRasmus Larsen is a Distinguished Engineer based in San Jose with over two decades of experience spanning numerical linear algebra, inverse problems, machine learning, and high-performance software for hardware accelerators. After a long tenure at Google Research where he advanced ML compilers and buffer assignment in XLA, he now leads linear algebra library work at NVIDIA, bridging algorithmic theory with production-grade performance engineering. His open-source contributions to projects like LFortran, Eigen, and XLA reflect deep expertise in compiler intrinsics, tensor operations, and memory liveness analysis that improve both correctness and speed. Trained as a PhD computer scientist with postdoctoral work at Stanford on inverse problems in helioseismology, he combines academic rigor with practical systems delivery. Colleagues describe him as someone who surfaces subtle numerical and performance pitfalls early—often by refactoring or renaming core primitives to avoid cross-library collisions—making complex algorithms reliably fast on modern hardware.
10 years of coding experience
25 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Aarhus Universitypost doc Computer Science / Solar physics, post doc Computer Science / Solar physics at Stanford University
Danish, English, German, French
Stackoverflow
44reputation
2kreached
5answers
0questions
Github Skills (39)
javascript10
visualization10
c-language10
matrix10
operation10
tensorrt10
fortran10
mat10
compiler-compiler10
tensorflow10
front-end-development10
hla10
xla10
compiler10
tensor10
Programming languages (11)
Github contributions (5)
eigenteam/eigen-git-mirror
May 2018 - Dec 2019
THIS MIRROR IS DEPRECATED -- New url: https://gitlab.com/libeigen/eigen
Role in this project:
Back-end Developer
Back-end DeveloperContributions:231 commits in 1 year 6 months
Contributions summary:Rasmus's commits primarily involved implementing and refining features within the Eigen library, specifically related to tensor operations. They developed vectorized clip functors for Eigen Tensors, enabling efficient clipping operations. Furthermore, the user refactored code to use `numext::maxi` and `numext::mini`, preventing collisions and implementing a rename from `scalar_clip_op` to `scalar_clamp_op` to prevent conflicts with other libraries. Additional contributions were focused on correcting potential performance bottlenecks in the library.
lapackeigenvalueslinear-algebrasparse-matrixeigen
openxla/xla
Jan 2017 - Apr 2018
A machine learning compiler for GPUs, CPUs, and ML accelerators
Role in this project:Backend Developer
Backend DeveloperContributions:29 commits in 1 year 2 months
Contributions summary:Rasmus primarily made changes to the buffer assignment, specifically in the XLA (Accelerated Linear Algebra) service. These changes involved modifications to buffer allocation, buffer assignment for computations, and related code related to the liveness analysis of buffers. Further commits reveal internal modifications to the test suite, particularly within the XLA tests. The user also contributed to modifying fast math flags and fixing errors in the shape utility within the XLA service.
compilermachine-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.