Zongwei Zhou
Software Engineer at Meta
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolZongwei Zhou is a software engineer and ML/AI performance expert with 10 years of experience building high-performance systems from hardware-aware compilers to cloud-native runtimes. He has led ML performance efforts at Google—optimizing XLA, TPUs, and large-model workflows—and now contributes to advanced research at Meta’s Superintelligence Lab. As a founding engineer at a startup he designed runtime libraries that boosted throughput up to 4x and halved latency across diverse databases and web stacks, and earlier work at VMware produced NVM-optimized in-memory stores and multiple patents. Zongwei is also an active contributor to TensorFlow models, adding mixed-precision Transformer optimizations and benchmarks for ResNet50 and BERT, reflecting a practical focus on measurable ML speedups. He combines deep academic training (PhD-level work at Carnegie Mellon) with hands-on systems engineering and hardware-software co-design, making him effective at turning cutting-edge research into production gains. Based in Mountain View, he brings a rare blend of compiler, hardware, distributed systems, and security background to large-scale AI infrastructure challenges.
10 years of coding experience
3 years of employment as a software developer
Master of Computer Science Computer Science, Master of Computer Science Computer Science at Tsinghua UniversityDoctor of Philosophy (Ph.D.) Computer Engineering, Doctor of Philosophy (Ph.D.) Computer Engineering at Carnegie Mellon University
Chinese, English, Mandarin
Github Skills (10)
machine-learning10
benchmarking10
benchmark10
deep-learning10
tensorflow10
python10
bert9
resnet9
eager-execution9
multiprecision8
Programming languages (4)
Github contributions (5)
tensorflow/models
Jul 2019 - Oct 2020
Models and examples built with TensorFlow
Role in this project:
ML Engineer
ML EngineerContributions:45 commits, 7 PRs, 20 pushes in 1 year 3 months
Contributions summary:Zongwei's contributions centered on enhancing the performance and functionality of TensorFlow models within the repository. Their work included optimizing the softmax function for mixed-precision training in the Transformer model, adding benchmarks for Resnet50 using eager execution and distribution strategies, and introducing benchmarks for BERT SQuAD. They also made internal changes within the vision image classification.
deep-learningtensorflow
zongweiz/tachyon
Aug 2015 - Aug 2015
Contributions:6 pushes, 1 branch in 1 day
memorydistributed-storagereliablestoragedistributed
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.