A Domain-Specific Supercomputer for Training Deep Neural Networks: Google's TPU supercomputers train deep neural networks 50x faster than general-purpose supercomputers running a high-performance computing benchmark.

The article examines the construction of a supercomputer by the high technology firm Google, featuring the first production domain specific architecture (DSA) for training deep neural networks (DNNs).

Detalles Bibliográficos
Publicado en:Communications of the ACM Vol. 63; no. 7; pp. 67 - 79
Autores principales: JOUPPI, NORMAN P., DOE HYUN YOON, KURIAN, GEORGE, SHENG LI, PATIL, NISHANT, LAUDON, JAMES, CLIFF YOUNG, PATTERSON, DAVID
Formato: Artículo
Publicado: Association for Computing Machinery Jul2020
Materias:
Acceso en línea:Ver este registro en EBSCOhost