Lance Wang
Tech Lead, Software Engineer at Google
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolLance Wang is a Tech Lead and software engineer with 10+ years of experience specializing in LLM training and serving on GPU/TPU platforms, currently driving LLM productionization and cross-framework unification at Google. He has deep expertise in model and data parallelism, attention-kernel and MoE optimizations, and has enabled models like Llama, Mistral, Gemma and GPT-J across JAX, PyTorch and TensorFlow on Google’s OSS stack (notably contributing to Tunix). His work spans low-level kernel work to system-level serving features such as SPMD/MPMD, continuous batching, disaggregated serving and speculative decoding, making him fluent across research-to-production ML pipelines. Prior roles at Microsoft and Marvell ground his background in systems, verification and hardware-aware design, and he holds advanced degrees in VLSI and quantitative analysis/CE from University of Cincinnati.
2 years of coding experience
9 years of employment as a software developer
MS, VLSI, MS, VLSI at University of CincinnatiBachelor, Telecommunication, Bachelor, Telecommunication at Southeast University
Github Skills (21)
pytorch10
batching10
large-language-models10
llm10
xla10
gpt10
inference9
deep-learning9
compiler8
tpu8
llm-inference7
llama7
gpu7
gemma7
transformer6
Programming languages (2)
Github contributions (5)
AI-Hypercomputer/jetstream-pytorch
Apr 2024 - Dec 2024
PyTorch/XLA integration with JetStream (https://github.com/google/JetStream) for LLM inference"
Contributions:1 release, 69 reviews, 37 PRs in 7 months
batchinggemmainferencellamallama2
wang2yn84/xla
Feb 2024 - Feb 2024
Enabling PyTorch on XLA Devices (e.g. Google TPU)
Contributions:2 pushes in 1 day
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.