Summary
Yuchung Cheng is a Director of Engineering at Google with 15 years of experience specializing in TCP, Linux kernel networking, and large-scale datacenter and Internet protocols. He leads host networking efforts for ML infrastructure and Google Cloud, focusing on optimizing stacks for LLM TPU training, availability, load balancing, and production RPC systems like Stubby/gRPC. Previously a Senior Staff Engineer, he co-invented and upstreamed influential TCP features such as BBR, RACK, TCP Fast Open, Tail Loss Probe, and PRR that shape modern congestion control. His background blends deep research (Ph.D. from UC San Diego) with production impact across Google’s networking infrastructure, bridging protocol design and system engineering. Based in Mountain View, he combines academic rigor with hands-on kernel and cloud optimizations, often re-architecting host networking for extreme-scale ML workloads. An often-overlooked strength is his track record of turning protocol research into deployable Linux and cloud primitives that directly improve global network performance.
15 years of coding experience
20 years of employment as a software developer
University of California San Diego
BS, Computer Science, BS, Computer Science at National Taiwan University