Summary
Kan Wu is a Member of Technical Staff focused on machine learning systems, caching, and distributed systems with a decade of experience bridging research and production. He earned a PhD in Computer Science from UW–Madison and has driven low-level serving and fleet-efficiency work at Google—optimizing Gemini serving, activation sparsity, memory tiering, and TLB efficiency—before joining xAI to work on inference for AGI. His background includes academic research at Microsoft Gray System Lab and early systems work at VMware and USTC, giving him a strong blend of systems research and engineering. Notably, he specializes in squeezing efficiency from compute and memory stacks to make large-model inference practical at scale.
10 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy (PhD) Computer Sciences, Doctor of Philosophy (PhD) Computer Sciences at University of Wisconsin-Madison
Bachelor of Engineering (B.Eng.) Computer Science and Technology, Bachelor of Engineering (B.Eng.) Computer Science and Technology at University of Science and Technology of China
Chinese, English