Motivation for and Evaluation of the First Tensor Processing Unit
Norman Paul Jouppi, Cliff Young, Nishant P. Patil, David A. Patterson · IEEE Micro · 2018
The first-generation tensor processing unit (TPU) runs deep neural network (DNN) inference 15-30 times faster with 30-80 times better energy efficiency than contemporary CPUs and GPUs in similar semiconductor technologies. This domain-specific architecture (DSA) is a custom chip that has been deployed in Google datacenters since 2015, where it serves billions of people.