Software Research Engineer - High Performance Computing And Embedded Systems Software at Lycée Chaptal
France
Join Prog.AI to see contacts
Join Prog.AI to see contacts
Summary
👤
Senior
🎓
Top School
Pierre Kestener is a Software Research Engineer with 11+ years of experience specializing in high-performance computing and embedded systems, focused on porting and optimizing numerical simulation codes (CFD, DFT, LBM) for large GPU and multicore clusters. He combines deep GPU programming expertise (CUDA, OpenCL, SYCL, Kokkos/C++) with MPI/OpenMP parallelism and pragmatic build-system work, having contributed to notable projects like Kokkos and RamsesGPU. At CEA he leads refactoring efforts to make scientific solvers architecture- and performance-portable, and also brings experience in real-time embedded software for space missions. An educator and trainer, he teaches GPU programming and PINNs and maintains open-source tooling (xsmurf, CanoP) linking research, reproducibility, and production HPC. Notably, his background in image processing and a PhD in electrical engineering underpin a rare blend of algorithmic insight and low-level systems craftsmanship.
11 years of coding experience
9 years of employment as a software developer
agrégation de physique - option physique appliquée, Electronique, Electrotechnique et Automatique, agrégation de physique - option physique appliquée, Electronique, Electrotechnique et Automatique at École normale supérieure Paris-Saclay
Doctor of Philosophy - PhD, Electrical engineering and image processing, Doctor of Philosophy - PhD, Electrical engineering and image processing at Université Bordeaux I
CPGE, Maths-Physique-Chimie - Spé P', CPGE, Maths-Physique-Chimie - Spé P' at Lycée Henri Poincaré, Nancy
Kokkos C++ Performance Portability Programming Ecosystem: The Programming Model - Parallel Execution and Memory Abstraction
Role in this project:
Automation Engineer / Build & Release Engineer
Contributions:2 reviews, 17 commits, 8 PRs in 3 years 1 month
Contributions summary:Pierre primarily contributed to build system and example configurations within the Kokkos repository. Their work involved modifying `generate_makefile.bash` to include new example builds and updating `Dependencies.cmake` to integrate examples within the Trilinos CMake build system. Furthermore, they addressed an issue with missing OpenMP linking flags in `kokkos_build.cmake`, demonstrating a focus on ensuring correct build configurations and dependency management for the project.
This repository consists for gpu bootcamp material for HPC and AI
Role in this project:
ML Engineer
Contributions:5 commits, 2 PRs in 4 months
Contributions summary:Pierre primarily contributed to the repository by addressing typos and inconsistencies within the project's documentation, specifically related to mathematical notation within the notebooks. They also made modifications to code files related to machine learning models, modifying neural network architectures and adapting code for cluster execution. These changes indicate a focus on refining the models and ensuring they are correctly implemented and executed in the intended environment.
cudampiopenaccdata-sciencedeep-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.