In-Datacenter Performance Analysis of a Tensor Processing Unit

P. Jouppi, Norman; Young, Cliff; Patil, Nishant; A. Patterson, David; Agrawal, Gaurav; Bajwa, Raminder; C. Bates, S.; Bhatia, Suresh; Boden, Nan; Borchers, Al; Boyle, Rick; Cantin, Pierre-luc; Chao, Clifford; Clark, Chris; Coriell, Jeremy; Daley, Mike; Dau, Matt; B. Dean, Jay; Gelb, Ben; Vazir Ghaemmaghami, Tara; Gottipati, Rajendra; Gulland, William; B. Hagmann, Robert; Richard Ho, C.; Hogberg, Doug; Wei-Shan Hu, John; Hundt, Robert; Hurt, Dan; Ibarz, Julian; Jaffey, Aaron; Jaworski, Alek; Kaplan, Alexander; Khaitan, Harshit; Killebrew, Daniel; Koch, Andy; Kumar, Naveen; Lacy, Steve; Laudon, James; Law, James; Le, Diemthu; Leary, Chris; Liu, Zhuyuan; Lucke, Kyle; Lundin, Alan; MacKean, Gordon; Maggiore, Adriana; Mahony, Maire; Miller, Kieran; Nagarajan, Rahul; Narayanaswami, Ravi

doi:10.1145/3079856.3080246

Public

In-Datacenter Performance Analysis of a Tensor Processing Unit

Shared by NobleBlocks on Jun 15, 2017 • 12:00 AM UTC

Authors:

Norman P. Jouppi

Cliff Young

Nishant Patil

Abstract

Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU) --- deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The ...

Subject

Computer science

Central processing unit

Application-specific integrated circuit

Finding related papers...

Discussions

(0)

No comments yet

Be the first to share your thoughts!