Performance of convolution neural network based on multiple GPUs with different data communication models

Che Lun Hung, Yi Yang Lin, Chuan Yi Tang, Chilung Wang, Ming Chiang Chen

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

1 Scopus citations

Abstract

Recently, deep learning technologies have been utilized in many scientific domains successfully. Convolution neural networks are common used in image understanding problems. However, to train a convolution neural network model with huge amount of images is time-consuming task. Most of deep learning frameworks, such as Caffe, TensorFlow, Torch, Keras, MxNet, and so forth, support GPU to train model fast; especially executing these models on multiple GPUs. In this work, we present the comparison of computation performance of AlexNet among different GPU servers and hyperparameters. The results shows that GPU servers with high bandwidth rate, NVLINK, can achieve better performance than others.

Original languageEnglish
Title of host publicationProceedings - 2018 IEEE/ACIS 19th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2018
EditorsHa Jin Hwang, Lizhi Cai, Gun Huck Yeom, Tokuro Matsuo, Haeng Kon Kim, Hyun Yeo, Chung Sun Hong, Naoki Fukuta, Takayuki Ito, Huaikou Miao
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages87-92
Number of pages6
ISBN (Print)9781538658895
DOIs
StatePublished - 20 Aug 2018
Event19th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2018 - Busan, Korea, Republic of
Duration: 27 Jun 201829 Jun 2018

Publication series

NameProceedings - 2018 IEEE/ACIS 19th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2018

Conference

Conference19th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2018
Country/TerritoryKorea, Republic of
CityBusan
Period27/06/1829/06/18

Keywords

  • Convolution Neural Networks
  • Deep Learning
  • GPU
  • Multiple GPUs

Fingerprint

Dive into the research topics of 'Performance of convolution neural network based on multiple GPUs with different data communication models'. Together they form a unique fingerprint.

Cite this