Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms

Young Jong Mo; Joongheon Kim; Jong Kook Kim; Aziz Mohaisen; Woojoo Lee

doi:10.1109/ICUFN.2017.7993784

Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms

Young Jong Mo, Joongheon Kim, Jong Kook Kim, Aziz Mohaisen, Woojoo Lee

School of Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

18 Citations (Scopus)

Abstract

In this paper we measure and verify the performance improvements in deep learning computation under the support of GPU-enabled multi-core parallel computing platforms. To measure the performance practically, we built our own computing platforms using a GPU hardware (1152 cores) and the TensorFlow software library. In order to evaluate the performance with GPU, we conducted the deep learning computation with various numbers of hidden layers in multilayer perceptron. As presented in the comparative performance results, utilizing GPU hardware improved the performance in terms of computation time (about 3 times or even more).

Original language	English
Title of host publication	ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks
Publisher	IEEE Computer Society
Pages	240-242
Number of pages	3
ISBN (Electronic)	9781509047499
DOIs	https://doi.org/10.1109/ICUFN.2017.7993784
Publication status	Published - 2017 Jul 26
Event	9th International Conference on Ubiquitous and Future Networks, ICUFN 2017 - Milan, Italy Duration: 2017 Jul 4 → 2017 Jul 7

Publication series

Name	International Conference on Ubiquitous and Future Networks, ICUFN
ISSN (Print)	2165-8528
ISSN (Electronic)	2165-8536

Other

Other	9th International Conference on Ubiquitous and Future Networks, ICUFN 2017
Country/Territory	Italy
City	Milan
Period	17/7/4 → 17/7/7

ASJC Scopus subject areas

Computer Networks and Communications
Computer Science Applications
Hardware and Architecture

Access to Document

10.1109/ICUFN.2017.7993784

Cite this

Mo, Y. J., Kim, J., Kim, J. K., Mohaisen, A., & Lee, W. (2017). Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms. In ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks (pp. 240-242). [7993784] (International Conference on Ubiquitous and Future Networks, ICUFN). IEEE Computer Society. https://doi.org/10.1109/ICUFN.2017.7993784

Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms. / Mo, Young Jong; Kim, Joongheon; Kim, Jong Kook et al.

ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks. IEEE Computer Society, 2017. p. 240-242 7993784 (International Conference on Ubiquitous and Future Networks, ICUFN).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Mo, YJ, Kim, J, Kim, JK, Mohaisen, A & Lee, W 2017, Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms. in ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks., 7993784, International Conference on Ubiquitous and Future Networks, ICUFN, IEEE Computer Society, pp. 240-242, 9th International Conference on Ubiquitous and Future Networks, ICUFN 2017, Milan, Italy, 17/7/4. https://doi.org/10.1109/ICUFN.2017.7993784

Mo YJ, Kim J, Kim JK, Mohaisen A, Lee W. Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms. In ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks. IEEE Computer Society. 2017. p. 240-242. 7993784. (International Conference on Ubiquitous and Future Networks, ICUFN). https://doi.org/10.1109/ICUFN.2017.7993784

@inproceedings{39e1e8009db44348a469798ea7abb688,

title = "Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms",

abstract = "In this paper we measure and verify the performance improvements in deep learning computation under the support of GPU-enabled multi-core parallel computing platforms. To measure the performance practically, we built our own computing platforms using a GPU hardware (1152 cores) and the TensorFlow software library. In order to evaluate the performance with GPU, we conducted the deep learning computation with various numbers of hidden layers in multilayer perceptron. As presented in the comparative performance results, utilizing GPU hardware improved the performance in terms of computation time (about 3 times or even more).",

author = "Mo, {Young Jong} and Joongheon Kim and Kim, {Jong Kook} and Aziz Mohaisen and Woojoo Lee",

note = "Funding Information: ACKNOWLEDGEMENT Joongheon Kim is a corresponding author of this paper. This work was supported by National Research Foundation of Korea (NRF Korea) under Grant 2016R1C1B1015406. Publisher Copyright: {\textcopyright} 2017 IEEE.; 9th International Conference on Ubiquitous and Future Networks, ICUFN 2017 ; Conference date: 04-07-2017 Through 07-07-2017",

year = "2017",

month = jul,

day = "26",

doi = "10.1109/ICUFN.2017.7993784",

language = "English",

series = "International Conference on Ubiquitous and Future Networks, ICUFN",

publisher = "IEEE Computer Society",

pages = "240--242",

booktitle = "ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks",

}

TY - GEN

T1 - Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms

AU - Mo, Young Jong

AU - Kim, Joongheon

AU - Kim, Jong Kook

AU - Mohaisen, Aziz

AU - Lee, Woojoo

N1 - Funding Information: ACKNOWLEDGEMENT Joongheon Kim is a corresponding author of this paper. This work was supported by National Research Foundation of Korea (NRF Korea) under Grant 2016R1C1B1015406. Publisher Copyright: © 2017 IEEE.

PY - 2017/7/26

Y1 - 2017/7/26

N2 - In this paper we measure and verify the performance improvements in deep learning computation under the support of GPU-enabled multi-core parallel computing platforms. To measure the performance practically, we built our own computing platforms using a GPU hardware (1152 cores) and the TensorFlow software library. In order to evaluate the performance with GPU, we conducted the deep learning computation with various numbers of hidden layers in multilayer perceptron. As presented in the comparative performance results, utilizing GPU hardware improved the performance in terms of computation time (about 3 times or even more).

AB - In this paper we measure and verify the performance improvements in deep learning computation under the support of GPU-enabled multi-core parallel computing platforms. To measure the performance practically, we built our own computing platforms using a GPU hardware (1152 cores) and the TensorFlow software library. In order to evaluate the performance with GPU, we conducted the deep learning computation with various numbers of hidden layers in multilayer perceptron. As presented in the comparative performance results, utilizing GPU hardware improved the performance in terms of computation time (about 3 times or even more).

UR - http://www.scopus.com/inward/record.url?scp=85028075671&partnerID=8YFLogxK

U2 - 10.1109/ICUFN.2017.7993784

DO - 10.1109/ICUFN.2017.7993784

M3 - Conference contribution

AN - SCOPUS:85028075671

T3 - International Conference on Ubiquitous and Future Networks, ICUFN

SP - 240

EP - 242

BT - ICUFN 2017 - 9th International Conference on Ubiquitous and Future Networks

PB - IEEE Computer Society

T2 - 9th International Conference on Ubiquitous and Future Networks, ICUFN 2017

Y2 - 4 July 2017 through 7 July 2017

ER -

Performance of deep learning computation with TensorFlow software library in GPU-capable multi-core computing platforms

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this