Asynchronous SGD for DNN training on shared-memory parallel architectures

Florent Lopez, Edmond Chow, Stanimire Tomov, Jack Dongarra

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

7 Scopus citations

Abstract

We present a parallel asynchronous Stochastic Gradient Descent algorithm for shared memory architectures. Different from previous asynchronous algorithms, we consider the case where the gradient updates are not particularly sparse. In the context of the MagmaDNN framework, we compare the parallel efficiency of the asynchronous implementation with that of the traditional synchronous implementation. Tests are performed for training deep neural networks on multicore CPUs and GPU devices.

Original languageEnglish
Title of host publicationProceedings - 2020 IEEE 34th International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages995-998
Number of pages4
ISBN (Electronic)9781728174457
DOIs
StatePublished - May 2020
Event34th IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020 - New Orleans, United States
Duration: May 18 2020May 22 2020

Publication series

NameProceedings - 2020 IEEE 34th International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020

Conference

Conference34th IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020
Country/TerritoryUnited States
CityNew Orleans
Period05/18/2005/22/20

Keywords

  • Asynchronous iterative methods
  • Deep learning
  • GPU
  • Multicore CPU
  • Stochastic Gradient Descent

Fingerprint

Dive into the research topics of 'Asynchronous SGD for DNN training on shared-memory parallel architectures'. Together they form a unique fingerprint.

Cite this