Paul Fultz is a seasoned software engineer with 14 years of experience building high-performance C++ systems and GPU-focused computational libraries, currently based in Austin and working at AMD. He has deep hands-on expertise in GPU tooling and testing—contributing to notable ROCm projects like HIP and Tensile by improving test infrastructure, multithreaded validation, and backend automation to make complex GPU benchmarks more runnable and robust. His background spans full-stack and embedded work, from web front-ends and deployment automation at Cisco to embedded web servers and cross-compilation toolchains, reflecting a pragmatic ability to bridge low-level performance engineering with higher-level application concerns. Paul has contributed to established open-source tooling such as cppcheck, showing a sustained focus on code quality, refactoring, and static analysis improvements. Colleagues would describe him as a developer who pairs thorough testing and automation with performance-minded design, able to debug integration issues across CUDA/ROCm environments. He brings a mix of research-collaboration experience and production software delivery, often making projects easier to build and validate without heavy installation overhead.
14 years of coding experience
7 years of employment as a software developer
Bachelor's Degree, Computer Science, 3.1, Bachelor's Degree, Computer Science, 3.1 at Wright State University
Contributions:484 reviews, 870 commits, 873 PRs in 4 years 10 months
Contributions summary:Paul made several commits related to the C/C++ code analysis tool cppcheck. Their work included adding features to classInfo, updating text and warnings, merging branches, refactoring code for performance, and adding options. The user demonstrated skills in modifying the core functionality of a C++ code analysis tool, which suggests a focus on back-end development and improving code quality.
Stretching GPU performance for GEMMs and tensor contractions.
Role in this project:
Backend & Automation Engineer
Contributions:2 reviews, 65 commits, 14 PRs in 3 years 5 months
Contributions summary:Paul primarily focused on improving the CobaltGen driver within the Tensile repository. They added entry points for benchmarking and simplified arguments to target the build process. Significant effort was put into making the project runnable without installation and allowing for the definition of CMake variables. Further contributions included creating a function for a Cobalt library, and setting up the necessary tools.
amdpythontensorhipauto-tuning
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.