Ruqing Xu
Deep Learning Engineer at NVIDIA
California, United States
Summary
🤩
Rockstar🎓
Top SchoolRuqing Xu is a Deep Learning Engineer at NVIDIA with eight years of experience specializing in performance optimization for numerical linear algebra and tensor computations. With a PhD in Physics from the University of Tokyo and a background in theoretical and mathematical physics, Ruqing blends rigorous scientific training with practical systems engineering. Their open-source work includes platform-aware build and release engineering for JuliaPackaging/Yggdrasil and high-performance BLIS kernel development for ARM SVE, demonstrating deep expertise in compiler/toolchain quirks and low-level GEMM optimization. Based in California, they focus on squeezing performance from both software stacks and hardware features, often tackling cross-platform build complexities that are easy to overlook. Pragmatic and detail-oriented, Ruqing bridges research-grade numerical methods and production-grade engineering to deliver robust, high-throughput deep learning primitives.
8 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, Physics, Doctor of Philosophy - PhD, Physics at 日本東京大學University of TokyoBachelor of Science - BS, Theoretical and Mathematical Physics, Bachelor of Science - BS, Theoretical and Mathematical Physics at University of Science and Technology of China
Github Skills (26)
assembly10
crossbuild10
matrix-multiplication10
build-system10
c-programming10
blas10
vector10
pbuilder10
sve10
performance-optimization10
assembler10
arm10
cross-compiling10
scalable10
linear-algebra10
Programming languages (9)
Github contributions (5)
JuliaPackaging/Yggdrasil
Sep 2020 - Apr 2022
Collection of builder repositories for BinaryBuilder.jl
Role in this project:
Backend Developer / Build & Release Engineer
Backend Developer / Build & Release EngineerContributions:13 reviews, 12 commits, 14 PRs in 1 year 7 months
Contributions summary:Ruqing primarily contributed to the build process and configuration of the `blis` and `tblis` libraries, which are crucial for numerical computing. Their work involved modifying build scripts, adapting configurations for different platforms (Linux, Windows, macOS, FreeBSD), and integrating with the BinaryBuilder.jl framework. They addressed compiler incompatibilities, optimized builds, and incorporated upstream fixes, demonstrating expertise in build systems and platform-specific adaptations within the context of numerical libraries. Additionally, the user also added the build recipe for a new dependency, Pfapack.
repositoriesbinarybuildermonorepobuilderjulia
flame/blis
Mar 2021 - Aug 2022
BLAS-like Library Instantiation Software Framework
Role in this project:Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:7 reviews, 69 commits, 17 PRs in 1 year 6 months
Contributions summary:Ruqing primarily focused on optimizing and extending the BLIS library's functionality for the Arm architecture. Their contributions include fixing bugs in the typed API definitions and API documentation. A significant portion of their work involved implementing new Arm SVE-based and row-major GEMM kernels, as well as improving the performance of existing ones. Furthermore, the user modified existing packing kernels for the ARM architecture.
software-frameworklinear-algebra-librarylinear-algebramatrix-multiplicationblis
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.