In this video from the Intel HPC Developer Conference, Narayanan Sundaram from Intel presents: Deep Learning at 15 Petaflops.
“We present the first 15-PetaFLOP Deep Learning system for solving supervised and semi-supervised scientific pattern classification problems, optimized for Intel® Xeon Phi™. We use a hybrid of synchronous and asynchronous training to scale to ~9600 nodes of Cori on CNN and autoencoder networks.”