Abstract

Accelerating Deep Learning Applications (Presented by Mellanox Technologies)
Come join us and learn how to build a data-centric GPC cluster for artificial intelligence. Mellanox is a leader in hig-performance, scalable, low-latency network interconnects for bot InfiniBand and Eternet. We will briefly present the state of the art techiques for distributed machine learning and what special requirements they impose on the system,
 


Session No: SIL8145
Speaker: Gil Bloch
Type:

Sponsored Sessions and NVIDIA Talks

Date: Thursday - October 18, 2018 03:00 PM - 03:25 PM
Location: Hall H
Topic: Accelerated Data Science
Industry: Aerospace

Come join us and learn how to build a data-centric GPC cluster for artificial intelligence. Mellanox is a leader in high-performance, scalable, low-latency network interconnects for both InfiniBand and Ethernet. We will briefly present the state of the art techniques for distributed machine learning and what special requirements they impose on the system, followed by an overview of interconnect technologies used to scale and accelerate distributed machine learning including RDMA, NVIDIA's GPCDirect technology, and in-network computing used to accelerate large scale deployments in HPC and artificial intelligence.