Google's Machine Learning Chip Is Up To 30x Faster, 80x More Efficient Than CPUs And GPUs

http://ift.tt/2nZ1qwj

Google revealed more details about its first machine learning chip, the Tensor Processing Unit (TPU). According to Google, the chip has 15-30x higher inference performance than a Haswell CPU and an Nvidia Tesla K80 GPU, and it is 40-80x more efficient.

from News Tom's Hardware http://ift.tt/2nLsz54
via IFTTT