Alexis Montoison
Membre Sénior De L'équipe Technique at AMD
Calgary, Alberta, Canada
Summary
🤩
Rockstar🎓
Top SchoolAlexis Montoison is a senior technical team member at AMD with eight years of experience building high-performance sparse linear algebra libraries for GPUs, currently working on rocSPARSE. Previously a postdoctoral researcher at Argonne National Laboratory, Alexis developed scalable linear solvers for exascale optimization codes, blending deep numerical analysis with production-focused software engineering. He contributes to prominent open-source projects in the Julia ecosystem—helping build BinaryBuilder recipes in Yggdrasil and adding CUDA-backed linear algebra and sparse operations to CUDA.jl—showing a rare combination of build-system expertise and GPU algorithm work. Comfortable patching complex cross-platform builds (including musl and macOS) and resolving OpenMP/CUDA compatibility issues, he bridges low-level performance tuning and reproducible packaging. Based in Calgary, his background includes a doctoral degree in mathematics and computer science from Polytechnique Montréal and engineering training from ENSEEIHT. Colleagues rely on him for pragmatic solutions where numerical stability, build portability, and GPU performance intersect.
8 years of coding experience
Doctor's Degree, Mathematics and Computer Science, Doctor's Degree, Mathematics and Computer Science at Polytechnique MontréalENSEEIHT
Github Skills (25)
freebuilder10
c-language10
gpu-programming10
pybuilder10
finalbuilder10
cmake10
app-builder10
compiler-options10
pbuilder10
jbuilder10
compile10
cuda10
compiler10
gui-builder10
ui-builder10
Programming languages (15)
Github contributions (5)
JuliaPackaging/Yggdrasil
Feb 2020 - Dec 2022
Collection of builder repositories for BinaryBuilder.jl
Role in this project:
Backend Developer
Backend DeveloperContributions:2 releases, 111 reviews, 64 commits in 2 years 10 months
Contributions summary:Alexis's primary contribution involves adding and refining builders for the NOMAD project, specifically focusing on building and compiling NOMAD with various compilers and configurations. The commits demonstrate a deep understanding of the build process, with the user adding patches, managing dependencies, and configuring build environments for different platforms, including those with musl and Apple's operating system. The work includes addressing OpenMP compilation issues and fixing build issues related to dependencies. Furthermore, there are contributions to the building of the PROPACK project.
repositoriesbinarybuildermonorepobuilderjulia
JuliaGPU/CUDA.jl
Sep 2020 - Dec 2022
CUDA programming in Julia.
Role in this project:Back-end Developer
Back-end DeveloperContributions:108 reviews, 23 commits, 90 PRs in 2 years 3 months
Contributions summary:Alexis contributed to the `cuda.jl` repository by implementing and extending functionality related to GPU-accelerated linear algebra operations within the Julia programming language. Their work involved adding new functions such as `swap!`, `rot!`, `rotate!`, and `reflect!` to the CUDA backend. They also added an option to the `sv2!` functions and implemented support for sparse matrix operations (sv!, sm!), which improved the library's coverage. Additionally, the user addressed an issue in the SpMV routines for CUDA 11.5, indicating a focus on performance and compatibility with different CUDA versions.
cudagpucuda-programmingjulia
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.