Zoy Bai
Datacenter GPU (MI250X Navi32) Staff Software Engineer
Austin, Texas, United States
Summary
👤
Senior🎓
Top SchoolZoy Bai is a staff software engineer focused on datacenter GPU and CPU performance for HPC and machine learning, currently driving MI250X/Navi32 work at AMD. With eight years of experience across AMD, Arm, and NXP, he blends low-level kernel and driver expertise (AMDGPU, Linux kernel, IOMMUs) with parallel programming and microarchitecture-aware optimization for x86-64 and arm64. He has shipped end-to-end performance work spanning ROCm/HIP, CUDA/OpenCL, UCX/MPI, NVMe/PCIe, and high-performance networking (InfiniBand/RoCE, DPDK), and previously optimized scalable server software stacks like TensorFlow.js, Nginx, and Memcached. Zoy’s background in CPU design and cache-coherence (Neoverse N1, CMN-600, CHI) gives him an uncommon systems-level perspective that bridges silicon, firmware, and distributed software. Based in Austin, he pairs rigorous academic training (M.S. in Computer Engineering from Texas A&M) with practical kernel-hacking and production performance tuning for exascale-class systems.
8 years of coding experience
5 years of employment as a software developer
Bachelor's degree, Electronic Information Engineering, 3.5, Bachelor's degree, Electronic Information Engineering, 3.5 at Beijing University of Posts and TelecommunicationsMaster's degree, Computer Engineering, 3.91, Master's degree, Computer Engineering, 3.91 at Texas A&M University
French, English, Chinese
Github Skills (21)
microbenchmark10
synchronization10
fasthttp8
server-sent-events7
lua6
pongo26
redis5
mysql4
postgresql4
tls134
sqlite4
go4
quic4
webserver4
react4
Programming languages (2)
Github contributions (5)
ARM-software/synchronization-benchmarks
May 2018 - Nov 2019
Collection of synchronization micro-benchmarks and traces from infrastructure applications
Contributions:81 commits, 48 PRs, 31 pushes in 1 year 6 months
microbenchmarkmicro-benchmarkssynchronization
zoybai/synchronization-benchmarks
Apr 2018 - Mar 2019
Collection of synchronization micro-benchmarks and traces from infrastructure applications
Contributions:12 PRs, 37 pushes, 25 branches in 10 months
tracesmicro-benchmarkssynchronizationinfrastructurebenchmarks
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.