Deep Learning Engineer @ Restb.ai
We need to realize that the current public dialog on AI — which focuses on a narrow subset of industry and a narrow subset of academia — risks blinding us to the challenges and opportunities that are presented by the full scope of AI, IA and II.
On ResNet-50, a single Cloud TPU (containing 8 cores and 64GB of RAM) is ~8.4 faster than a single P100 and ~5.1 times faster than a V100. For InceptionV3, the speedup is almost the same (~8.4 and ~4.8, respectively). With smaller precisi…
“Convolutional neural networks are doomed” — Geoffrey Hinton