News for hackers. Stuff that matters.
Re:Introducing Amazon EC2 P3 Instances by eggie5 on Thursday October 26, 2017 @05:47AM (#15557539)
Here's my results:

Testing new Tesla V100 on AWS. Fine-tuning VGG on DeepSent dataset for 10 epochs.

GRID 520K (4GB) (baseline):

* 780s/epoch @ minibatch 8 (GPU saturated)

V100(16Gb):

* 30s/epoch @ minibatch 8 (GPU not saturated)

* 6s/epoch @ minibatch 32 (GPU more saturated)

* 6s/epoch @ minibatch 256 (GPU saturated)

Re:Introducing Amazon EC2 P3 Instances by dharma1 on Thursday October 26, 2017 @08:36AM (#15558265)
Thanks! Curious how this would scale on the 8x or 16x instances
Re:Introducing Amazon EC2 P3 Instances by eggie5 on Thursday October 26, 2017 @05:45PM (#15563349)
what do you mean? 8 or 16 GPUs? That's require changing the code to use distributed tensorflow...
Re:Introducing Amazon EC2 P3 Instances by dharma1 on Thursday October 26, 2017 @06:02PM (#15563533)
Yes exactly. The instances with 8 or 16 GPUs. Does the training time reduce linearly, is the GPU utilisation 100%, is it plug and play with TF