Wenzel Jakob
Software Engineer at EPFL
Lausanne, Vaud, Switzerland
Summary
🤩
RockstarWenzel Jakob is a software engineer with 14 years of experience specializing in high-performance rendering, numerical algorithms, and native Python/C++ integration. Based in Lausanne, he contributes to and optimizes core components of prominent open-source projects such as Mitsuba, Enoki, CPython, and pybind11, with deep expertise in SIMD intrinsics, LLVM/CUDA backends and thread libraries like TBB. His work spans full-stack library maintenance to low-level performance engineering—e.g., adding AVX512/NEON intrinsics, JIT CUDA fixes, and vector call APIs—demonstrating a rare mix of graphics research pedigree and production-grade systems engineering. He also maintains GUI and build toolchains, improving cross-platform compatibility and developer ergonomics, and has a research profile in inverse graphics and material appearance modeling. Not obvious from titles: he frequently bridges research and tooling, turning physically based rendering ideas into maintainable, high-performance code used by both academic and industry projects.
15 years of coding experience
Stackoverflow
623reputation
31kreached
8answers
4questions
Github Skills (60)
cpython10
unit-testing10
algorithms10
mesh10
dependency-management10
algorithm10
develop10
c-language10
3d-graphics10
python10
llvm10
rendering10
avx10
raytracing10
build-system10
Programming languages (21)
Github contributions (5)
mitsuba-renderer/enoki
May 2016 - Sep 2020
Enoki: structured vectorization and differentiation on modern processor architectures
Role in this project:
Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:569 commits, 18 PRs, 135 pushes in 4 years 4 months
Contributions summary:Wenzel was primarily focused on enhancing the performance of the Enoki library through optimizations and extending its capabilities. Their work included the introduction of AVX512 and NEON intrinsics to accelerate key functions such as rcp(), rsqrt(), and shifted sampling, as well as refactoring for increased efficiency. Furthermore, the user contributed substantial changes related to a JIT-compiled CUDA backend, addressing memory allocation issues and enhancing code generation. Their commits also added support for specific types of vectors, making it clear they were working on the core performance-critical codebase.
vectorization
wjakob/tbb
Apr 2016 - Sep 2020
Intel TBB with CMake build system
Role in this project:Back-end Developer
Back-end DeveloperContributions:53 commits, 39 PRs, 138 pushes in 4 years 5 months
Contributions summary:Wenzel primarily contributes to updating the Intel TBB library. Their commits involve updating the TBB version, which includes modifications to documentation and build system files like CMake. The user also rebased the CMake build system on the current TBB version, showcasing involvement in the build process and dependency management. These updates and rebasings indicate a focus on maintaining and integrating the library into the project's build process.
build-systemcmake
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.