Dr. Matthias Kretz
Senior Software Engineer and SIMD/C++ Standardization Lead
Husband — Father — Nerd — Athlete
I am a Senior Software Engineer at GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt,
Germany, and chair of WG21 SG6 (Numerics).
With over a decade of contribution to the C++ standardization process, my work spans from contributions to KDE to
leading the development of std::simd for C++ and its implementation in GCC.
I aspire for efficiency by removing complexity—from making computers accessible through KDE to enabling scientists to write high-performance code without wrestling with vectorization details. I recognize that every abstraction adds its own complexity, but believe in shifting it to where it matters most: compile-time verification rather than runtime surprises.
My research evolved from using SIMD for track reconstruction at CERN to SIMD types (including my PhD work), leading to
my contribution of std::experimental::simd to the Parallelism TS and its eventual inclusion as std::simd into C++26.
I implement and maintain these in GCC, making my work accessible out-of-the box to most scientists.
I focus my language design efforts on std::simd and related numerics topics due to the specialized nature of this
work.
You can find all my WG21 papers by searching for “Kretz” at
open-std.org or explore the repository at
wg21-papers.
There is also my talk from
CppCon 2023: std::simd — How to Express Inherent Parallelism Efficiently via Data-Parallel
Types.
Research & Implementation Interests
- C++ Language Design — Consistency, orthogonality, high performance, simplicity, and general applicability
- Data Parallelism — SIMD, ILP, memory access patterns, performance portability, GPU computing
- Scientific Computing — Numerics support for HPC and High-Energy Physics workloads
- Compiler Optimization — Practical implementation experience driving “missed optimization” reports
Research Vision
Data-parallel types express data parallelism in general—they are not bound to CPUs with SIMD instructions. The same abstraction could translate to GPU execution or other parallel architectures. Achieving this requires novel compiler interactions to maintain the abstraction while exposing architecture-specific performance characteristics.
Selected Publications
- Kretz, M. (2015). Extending C++ for explicit data-parallel programming via SIMD vector types. Dissertation, Johann Wolfgang Goethe-Universität Frankfurt am Main. PDF. URN
- Heller, T., Adelstein Lelbach, B., Huck, K.A., et al. including Kretz, M. (2019). Harnessing billions of tasks for a scalable portable hydrodynamic simulation of the merger of two stars. The International Journal of High Performance Computing Applications, 33(4), 699-715. DOI
- Kretz, M., & Lindenstruth, V. (2011). Vc: A C++ library for explicit vectorization. Software: Practice and Experience. DOI
- Rohr, D., Gorbunov, S., Kretz, M., et al. (2012). ALICE HLT TPC Tracking of Pb-Pb Events on GPUs. Journal of Physics: Conference Series, 396(1). DOI
- Gorbunov, S., Rohr, D., Kretz, M., et al. (2011). ALICE HLT High Speed Tracking on GPU. IEEE Transactions on Nuclear Science, 58(4). DOI
- Kretz, M. (2009). Efficient Use of Multi- and Many-Core Systems with Vectorization and Multithreading. Diplomarbeit, Universität Heidelberg.
Honors & Awards
- Rainer-Kemp-Dissertations-Preis (2016) — Best dissertation in Computer Science and Mathematics, Goethe University Frankfurt
- Studienstiftung des deutschen Volkes (2001–2008) — Scholarship from Germany’s largest academic foundation
- KDE Akademy Award (2007) — “Best Non-Application” for Phonon multimedia framework
- Bundeswettbewerb Informatik (2001) — Winner of Germany’s National Computer Science Competition; Special Award for Cooperative Behavior
Software Projects
- std::simd Staging area before contribution to GCC
- vir-simd Extensions for
std::experimental::simd - C++ CI docker images
- value-preserving literals
- bit_int — Compile-time computation for bit-width integers
- Various WG21-related tools and patches
- See GitHub profile for complete list
- See Sourceware Forge for my GCC-related work
When I still had time for KDE:
- KDE Phonon library (KDE 4)
- KDE core library contributor for KDE 4
- aRts contributions (KDE 2/3 sound server)
- KView (KDE 2)
Connect
- Mastodon: @mkretz@floss.social
- GitHub: mattkretz
- ORCID: 0000-0002-0867-243X
- Garmin Connect: Profile
- Strava: Profile