Mark Philipp
Staff Software Engineer at NVIDIA
Fresno, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMark Philipp is a Staff Software Engineer based in Fresno, CA, specializing in backend systems and AI with five years of professional experience and a longer technical background in full-stack and tooling. He leads development of an AI-driven code review tool at NVIDIA, routinely working with frontier LLMs, prompt engineering, performance tuning, Kubernetes, databases, and large-scale infra. An active open-source contributor to ML compiler work, he has implemented cublasLT and FP8 GEMM support within XLA/TensorFlow to accelerate matrix math on modern GPU architectures. His earlier roles include building full-stack study management tools and Unity3D applications for research and educational customers, plus a decade of developer-facing content and support that drove significant community engagement. Mark blends research-grade ML systems work with practical product delivery and a knack for turning performance-focused research (FP8/GEMM optimizations) into production-ready compiler integrations.
5 years of coding experience
14 years of employment as a software developer
Computer Science, Computer Science at Clovis East High SchoolAssociate of Science (A.S.) Computer Science, Associate of Science (A.S.) Computer Science at Clovis Community CollegeMaster's of Science Computer Science - Computational Perception & Robotics, Master's of Science Computer Science - Computational Perception & Robotics at Georgia Institute of TechnologyComputer Science, Computer Science at Center for Advanced Research and TechnologyBachelor of Science - BS Computer Science, Bachelor of Science - BS Computer Science at California State University, Fresno
English
Github Skills (14)
cuda10
matrix-multiplication10
xla10
gpu-programming10
compiler10
machine-learning10
gemfire10
p810
c-language10
cublas10
tensorflow10
cprogramming-language10
cudnn10
deep-learning9
Programming languages (2)
Github contributions (5)
openxla/xla
Jan 2022 - Jan 2023
A machine learning compiler for GPUs, CPUs, and ML accelerators
Role in this project:
Back-end Developer
Back-end DeveloperContributions:75 reviews, 41 commits, 34 PRs in 1 year
Contributions summary:Mark's commits primarily focus on integrating `cublasLT` within the XLA framework, specifically targeting the `xla/service/gpu` directory. Their work involves modifying `gemm_thunk.cc`, `gemm_algorithm_picker.cc`, and related files to enable and leverage the `cublasLT` library for matrix multiplication operations, as well as related features such as vector bias and ReLU activation. The changes include implementing the necessary configurations and integrating the `cublasLT` APIs to optimize GEMM performance within XLA. The user also introduced a new test case for cublaslt.
compilercommunity-drivenmachine-learningmodular
tensorflow/tensorflow
May 2021 - Jan 2023
An Open Source Machine Learning Framework for Everyone
Role in this project:ML Engineer
ML EngineerContributions:137 reviews, 80 commits, 41 PRs in 1 year 8 months
Contributions summary:Mark contributed to the implementation of FP8 GEMMs (General Matrix Multiply) within the XLA (Accelerated Linear Algebra) framework, which is part of the larger TensorFlow project. Their work involved adding support for FP8 operations, specifically targeting Hopper architecture and newer systems, by modifying files related to the XLA compiler, tests, and math kernel libraries. The user appears to have focused on enabling and optimizing these operations, including adjusting the data layout and inserting a new custom call to handle the specific FP8 computations within cuDNN.
pythondata-sciencedeep-learningmlmachine-learning
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.