Fabian Ritter
Senior Software Development Engineer at AMD
Germany
Summary
🤩
Rockstar🎓
Top SchoolFabian Ritter is a Senior Software Development Engineer with a decade of experience focused on compiler backends and GPU toolchains, currently contributing to ROCm and the AMDGPU LLVM backend at AMD. He holds a summa cum laude Dr. Ing. in Computer Science from Saarland University and spent six years researching low-level CPU performance modeling and compiler-based memory safety while teaching compiler construction. An active open-source contributor to the widely used llvm-project, Fabian has implemented practical codegen optimizations for AMDGPU—improving memcpy/memmove handling, local memory load/store efficiency, and alignment in variable-sized copies. He blends academic rigor with production engineering, translating research insights into tangible compiler fixes that improve real-world GPU performance. Colleagues rely on him for deep systems expertise and careful, correctness-first improvements in complex compiler internals.
10 years of coding experience
Bachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at Universität des SaarlandesDr. Ing., Computer Science, summa cum laude, Dr. Ing., Computer Science, summa cum laude at Saarland University
German, English
Github Skills (11)
memory-management10
code-generation10
c-language10
amdgpu10
cprogramming-language10
llvm10
assemble9
assembler9
optimization9
assembly9
documentation7
Programming languages (6)
Github contributions (5)
llvm/llvm-project
Apr 2022 - Apr 2025
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Role in this project:
Backend Developer
Backend DeveloperContributions:107 reviews, 86 PRs, 101 pushes in 3 years
Contributions summary:Fabian primarily contributes to the LLVM project by addressing documentation issues, fixing broken links, and improving the code generation for AMDGPU. They implemented optimizations for memory copy (memcpy) operations, focusing on the efficient handling of load and store instructions, especially in the context of local memory. They also addressed the handling of volatile arguments in memmove intrinsics and corrected memory alignment issues in residual loops for variable-sized memory copies.
compilerstechnologiesclangsubmittoolchain
cdl-saarland/MemInstrument
Jul 2017 - May 2022
Contributions:314 commits in 4 years 10 months
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.