Training Neural Networks with Hundreds of GPUs on Graham and Cedar

September 20, 2017 @ 1:00 pm – 2:00 pm

Speaker: Fei Mao, SHARCNET

This webinar is a guide for doing very large scale Deep Learning research on the new systems. Both hardware and software capabilities will be introduced and discussed in details. Caffe2 and Tensorflow will be used as examples for training ResNet across multiple GPU nodes.

No registration is required. Join the seminar at SN-Seminars Vidyo room

Note: In order to receive audio/video you will need to install the Vidyo plugin (Windows, Mac, Linux, iPhone, Android), which should download automatically when you follow the above link. If you run into problems, please email It is highly recommended that you join the seminar a few minutes early to ensure Vidyo is working before the meeting starts.