CS Seminar Series: Chunqiang Tang

From Custom Silicon to Hyperscale AI Clusters: How Meta's Infrastructure Serves Billions of People Abstract: Each day, 3.6 billion people use Meta's products. Operating infrastructure at this scale requires progress on two fronts at once: foundational systems research and production engineering. In this talk, Dr. Tang gives an overview of how Meta's hyperscale infrastructure works in practice and describes the "research in production" philosophy that connects these two worlds. He then examines two areas in depth: (1) global resource pooling — abstracting computation across geographically distributed data centers into a single, cohesive pool of resources; and (2) AI cluster reliability — diagnosing hardware failures and silent data corruption at scale, and mitigating their impact on large-scale LLM training jobs. The talk closes with perspectives on where hyperscale AI infrastructure is heading, from silicon to systems, and on the emerging role of AI in recursive self-improvement. Bio: Dr. Chunqiang (CQ) Tang is Vice President of the AI & Compute Foundation organization at Meta, leading Meta's hardware strategy, hardware/software co-design, and AI chip software. Since joining Meta in 2013, he has helped build foundational infrastructure across the full hardware and software stack: hyperscale AI clusters, in-house silicon, AI chip software, LLM training and inference software, and internal cloud software. While his primary focus is on building production systems to serve billions of people, Dr. Tang is also an accomplished systems researcher. His work has earned Best Paper Awards at SOSP, OSDI, ISCA, and ASPLOS; multiple IEEE Micro Top Picks; and two Communications of the ACM honors—a Research Highlight and a cover story. He has published over 100 papers and holds around 100 patents. Dr. Tang holds a B.S. in Computer Science from the University of Science and Technology of China, an M.S. in Computer Science from Institute of Computing Technology, Chinese Academy of Sciences, and a Ph.D. in Computer Science from the University of Rochester.