Summary
Zi Yan is a software engineer with a decade of deep systems and research experience specializing in memory management, compilers, and computer architecture. Based in Greater Boston, he has driven production and upstream Linux kernel work at NVIDIA—shipping transparent huge page migration and memory defragmentation enhancements that measurably improved translation coverage. His background combines a PhD-level research foundation from Rutgers and hands-on GPU/LLM systems work (Megatron Core MoE, NVIDIA LLM training stacks), bridging low-level OS internals with large-scale ML infrastructure. Known for shipping kernel patches and performance-focused runtime features, he brings a rare blend of academic rigor and production-grade system engineering.
10 years of coding experience
SUN YAT-SEN UNIVERSITY