Shang Wang
Manager, AI Systems Software at NVIDIA
Toronto, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolShang Wang is a manager and hands-on AI systems software engineer with ~8 years of experience building and optimizing large-scale ML infrastructure, from CUDA/HIP kernels and PyTorch internals to Kubernetes-backed inference and training services. He co-founded CentML to productize those low-level optimizations and led projects at NVIDIA that cut GPU costs and accelerated BERT pretraining (e.g., LDDL) as well as contributing to TensorFlow XLA parsing and GPU command-buffer work. Academically grounded with an ongoing PhD and MSc from University of Toronto, he blends rigorous research with production engineering across industry leaders like NVIDIA, Google and Intel. Known for squeezing major performance wins out of complex systems, he also actively contributes to prominent open-source ML projects and benchmark infra (MLPerf). Based in Toronto, he hires and mentors engineers who enjoy deep systems problems and end-to-end ML performance engineering.
8 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of Toronto
Github Skills (8)
cuda10
machine-learning10
tensorflow10
gpu10
python9
deep-learning9
neural-network8
deep-neural-networks8
Programming languages (3)
Github contributions (5)
tensorflow/tensorflow
Dec 2022 - Dec 2022
An Open Source Machine Learning Framework for Everyone
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:14 reviews, 1 commit, 4 PRs in 1 day
Contributions summary:Shang contributed to the XLA project, focusing on enhancing the HLO parser to accommodate computations without module headers and improving tests related to parsing. Their work included modifying existing code files to incorporate features for parsing and handling computations, demonstrating a focus on the underlying workings of the machine learning framework. The user also made multiple changes related to GPU and CUDA, including adding command buffer support for different operations, implying involvement in optimizing and extending the framework for GPU usage.
pythondata-sciencedeep-learningmlmachine-learning
shawnwang18/xla
Apr 2023 - Mar 2025
A machine learning compiler for GPUs, CPUs, and ML accelerators
Contributions:351 pushes, 79 branches in 1 year 11 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.