“A Domain-Specific Supercomputer for Training Deep Neural Networks”, Norman P. Jouppi, Doe Hyun Yoon, George Kurian, Sheng Li, Nishant Patil, James Laudon, Cliff Young, David Patterson2020-06-01 (; similar)⁠:

Google’s TPU supercomputers train deep neural networks 50× faster than general-purpose supercomputers running a high-performance computing benchmark.

[cf. “The Design Process for Google’s Training Chips: TPUv2 and TPUv3”, Norrie et al 2021]

View PDF:

A Domain-Specific Supercomputer for Training Deep Neural Networks