Tobias Pietz
Software Engineer at scalable minds
Karlsruhe, Baden-Württemberg, Germany
Summary
🤩
Rockstar🎓
Top SchoolTobias Pietz is a software engineer with 14 years of experience, currently building scalable systems at scalable minds and pursuing research software work in numerical linear algebra at TUM. He combines deep low-level C++ and HPC expertise with practical build and packaging skills, evidenced by contributions to high-profile projects like NVIDIA/thrust, CMake, cuDF, vcpkg, ConanCenter and Spack. His work spans performance-critical algorithm design (e.g., unique_count in thrust and optimized multibyte_split kernels) to robust build/test integrations for CUDA and cross-platform packaging. Tobias has industrial experience from internships and projects at Google and NVIDIA, and a strong academic foundation from the Hasso Plattner Institute. Colleagues describe him as the kind of engineer who can both debug subtle parallel iterator issues and harden CI tooling to catch rare GPU errors.
14 years of coding experience
1 year of employment as a software developer
Gymnasium Walter GropiusMaster's degree, Master's degree at Hasso Plattner Institute
Github Skills (46)
dependency-manager10
dependency-management10
cpp-library10
thrust10
algorithms10
c-language10
gpgpu10
package-management10
dependency-analysis10
conan10
python10
testing10
cudf10
build-system10
cmake10
Programming languages (18)
Github contributions (5)
rapidsai/cudf
Aug 2022 - Oct 2022
cuDF - GPU DataFrame Library
Role in this project:
Back-end Developer
Back-end DeveloperContributions:198 reviews, 16 commits, 17 PRs in 2 months
Contributions summary:Tobias primarily focused on improving the `read_text` functionality of the cuDF library, addressing issues related to byte range alignment and field duplication when reading text data. They fixed bugs, added test cases to ensure correctness, and optimized the core `multibyte_split` kernel. The user implemented a special-case for single-byte delimiters, enhancing performance, and added a new BGZIP data chunk reader. Furthermore, the user refactored the `output_builder` class to improve modularity and implemented a strip_delimiters option.
cudadataframe-librarydata-analysiscppcudf
Kitware/CMake
May 2020 - Oct 2022
Mirror of CMake upstream repository
Role in this project:Back-end & QA Engineer
Back-end & QA EngineerContributions:12 commits in 2 years 4 months
Contributions summary:Tobias primarily contributed to the CMake project by adding support for CUDA with Clang and integrating cuda-memcheck. They modified existing CMake modules to incorporate CUDA compiler flags and implemented new features within CTest to parse and report cuda-memcheck results. Furthermore, the user added test cases to validate the parsing of different CUDA error types and integrated compute-sanitizer. This work directly enhances the testing capabilities for CUDA-related code within the CMake build system.
cppcmakeupstream
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.