David Patterson
Professor Emeritus of Computer Science at UC Berkeley
A computer-architecture pioneer whose later work on tensor-processing hardware helped connect efficient computing with modern machine learning.
Biography
David Patterson is an American computer scientist and professor emeritus at the University of California, Berkeley. He earned his doctorate at UCLA and spent decades investigating how computers can perform more useful work with efficient architectural designs. His research helped establish reduced instruction set computing and influenced generations of computer systems.
Patterson and John Hennessy received the 2017 ACM Turing Award for their contributions to computer architecture. Patterson’s significance to AI also includes work at Google on Tensor Processing Units, specialized hardware for neural-network computation. He co-authored research evaluating the first TPU’s behavior in real datacenter workloads.
His career shows how AI progress depends on computing design as well as learning algorithms. Specialized accelerators can change the cost and energy requirements of running neural networks, while systems research provides evidence about their practical performance. Patterson’s contribution should therefore be understood through architecture, measurement, and education, rather than as the invention of a particular AI model.
Notable contributions
Co-authored the first TPU performance study
Evaluated a specialized accelerator for neural-network inference in Google datacenters.
Received the ACM Turing Award with John Hennessy
Recognized for systematic contributions to computer architecture, including reduced instruction set computing.