Global ETD Search

Scalability Analysis and Optimization for Large-Scale Deep Learning

Pumma, SarunyaVirginia Tech2020

… degraded scalability. We propose various novel optimizations that minimize resource contention and improve performance by up to 35% for training various neural networks on 24,576 GPUs of the Summit supercomputer at Oak Ridge National Laboratory---the world's largest supercomputer at the time of …

vt Repository record for Scalability Analysis and Optimization for Large-Scale Deep Learning (opens in a new tab)