Phuong Nguyen
Sr. Performance Engineer, Deep Learning - Tech Lead at NVIDIA
California, United States
Summary
🤩
Rockstar🎓
Top SchoolPhuong Nguyen is a senior performance engineer and tech lead at NVIDIA with eight years of experience optimizing high-performance computing and deep learning software across CPUs and GPUs. She specializes in modern C++, CUDA and SYCL, and parallel programming frameworks (MPI, OpenMP), with a track record of porting and tuning batched solvers and kernels for diverse architectures including Intel and Arm-based GPUs. Her work bridges research and production: contributions at ICL and Intel include porting Ginkgo solvers to SYCL and achieving multi-fold speedups in Lattice QCD kernels through data-layout and cache-blocking optimizations. Based in California, she combines academic rigor from TUM and Stanford coursework with hands-on supercomputer and benchmarking experience from LRZ and CSC. Colleagues describe her as a pragmatic optimizer who uncovers non-obvious bottlenecks across memory hierarchies and heterogeneous systems to deliver measurable performance gains.
8 years of coding experience
5 years of employment as a software developer
Master's degree, Computational Science and Engineering, Master's degree, Computational Science and Engineering at Technical University of MunichBachelor's degree, Science and Technology - Specialized in Chemistry and Material Science, Bachelor's degree, Science and Technology - Specialized in Chemistry and Material Science at University of TartuSoftware Systems Graduate Certificate, Software Systems Graduate Certificate at Stanford University
English, Vietnamese
Github Skills (66)
python10
transformer-models10
hip10
deep-learning10
gpu10
floating10
floating-point10
transformer10
hopper10
jax10
utilization10
nvidia10
automatic-differentiation10
cuda10
8-bit10
Programming languages (7)
Github contributions (5)
phu0ngng/TransformerEngine
Apr 2024 - Apr 2025
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs, to provide better performance with lower memory utilization in both training and inference.
Contributions:17 PRs, 327 pushes, 99 branches in 11 months
phu0ngng/dotfiles
Nov 2022 - Mar 2025
Contributions:26 pushes in 2 years 5 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.