Guanhua Wang
Bellevue, Washington, United States
Summary
🤩
RockstarGuanhua Wang is a research scientist on Meta's PyTorch team specializing in accelerating GenAI training at production scale. He holds a PhD in Computer Science from UC Berkeley's AMPLab/RISELab under Professor Ion Stoica, with research focused on distributed machine learning and networking. Previously he was a senior researcher and tech lead on Microsoft's DeepSpeed team, driving projects like Domino and ZeRO++ to reduce communication overhead and speed large-scale training for Microsoft and OpenAI models, and working on GPU communication optimizations within DGX clusters. With about a decade of experience across research labs, industry internships, and consulting on distributed ML infrastructure, he combines academic rigor with systems-first engineering to make large-model training practical.
10 years of coding experience
Github Skills (90)
transformers10
cnn-classification10
python10
transformer-models10
deep-learning10
gpu10
compression10
smart-contracts10
gpu-acceleration10
efficientnet10
optimization10
distributed-training10
imagenet10
acceleration10
sysml10
Programming languages (5)
Github contributions (5)
GuanhuaWang/SystemPaper
Dec 2016 - Jun 2020
Contributions:313 commits, 311 pushes, 1 branch in 3 years 6 months
GuanhuaWang/sensAI
Mar 2019 - Jan 2021
sensAI: ConvNets Decomposition via Class Parallelism for Fast Inference on Live Data
Contributions:118 commits, 7 pushes in 1 year 11 months
cifar-10cifar-100cifar10cifar100cnn-classification
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.