On the energy benefits of spiking deep neural networks: A case study

Bing Han, Abhronil Sengupta, Kaushik Roy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

Deep learning neural networks have achieved success in a large number of visual processing tasks and are currently utilized for many real-world applications like image search and speech recognition among others. However, in spite of achieving high accuracy in such classification problems, they involve significant computational resources. Over the past few years, artificial neural network models have evolved into the biologically realistic and event-driven spiking neural networks. Recent research efforts have been directed at developing mechanisms to convert traditional deep artificial nets to spiking nets where the neurons communicate by means of spikes. However, there have been limited studies providing insights on the specific power, area and energy benefits offered by deep spiking neural nets in comparison to their non-spiking counterparts. In this paper, we perform a case study for a hardware implementation of a spiking/non-spiking deep net on the MNIST dataset and clearly outline the design prospects involved in implementing neural computing platforms in the spiking mode of operation.

Original languageEnglish (US)
Title of host publication2016 International Joint Conference on Neural Networks, IJCNN 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages971-976
Number of pages6
ISBN (Electronic)9781509006199
DOIs
StatePublished - Oct 31 2016
Event2016 International Joint Conference on Neural Networks, IJCNN 2016 - Vancouver, Canada
Duration: Jul 24 2016Jul 29 2016

Publication series

NameProceedings of the International Joint Conference on Neural Networks
Volume2016-October

Other

Other2016 International Joint Conference on Neural Networks, IJCNN 2016
CountryCanada
CityVancouver
Period7/24/167/29/16

Fingerprint

Neural networks
Speech recognition
Neurons
Hardware
Deep neural networks
Processing
Deep learning

All Science Journal Classification (ASJC) codes

  • Software
  • Artificial Intelligence

Cite this

Han, B., Sengupta, A., & Roy, K. (2016). On the energy benefits of spiking deep neural networks: A case study. In 2016 International Joint Conference on Neural Networks, IJCNN 2016 (pp. 971-976). [7727303] (Proceedings of the International Joint Conference on Neural Networks; Vol. 2016-October). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IJCNN.2016.7727303
Han, Bing ; Sengupta, Abhronil ; Roy, Kaushik. / On the energy benefits of spiking deep neural networks : A case study. 2016 International Joint Conference on Neural Networks, IJCNN 2016. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 971-976 (Proceedings of the International Joint Conference on Neural Networks).
@inproceedings{ba2a710fb8c2432a9edb6d271ccb7788,
title = "On the energy benefits of spiking deep neural networks: A case study",
abstract = "Deep learning neural networks have achieved success in a large number of visual processing tasks and are currently utilized for many real-world applications like image search and speech recognition among others. However, in spite of achieving high accuracy in such classification problems, they involve significant computational resources. Over the past few years, artificial neural network models have evolved into the biologically realistic and event-driven spiking neural networks. Recent research efforts have been directed at developing mechanisms to convert traditional deep artificial nets to spiking nets where the neurons communicate by means of spikes. However, there have been limited studies providing insights on the specific power, area and energy benefits offered by deep spiking neural nets in comparison to their non-spiking counterparts. In this paper, we perform a case study for a hardware implementation of a spiking/non-spiking deep net on the MNIST dataset and clearly outline the design prospects involved in implementing neural computing platforms in the spiking mode of operation.",
author = "Bing Han and Abhronil Sengupta and Kaushik Roy",
year = "2016",
month = "10",
day = "31",
doi = "10.1109/IJCNN.2016.7727303",
language = "English (US)",
series = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "971--976",
booktitle = "2016 International Joint Conference on Neural Networks, IJCNN 2016",
address = "United States",

}

Han, B, Sengupta, A & Roy, K 2016, On the energy benefits of spiking deep neural networks: A case study. in 2016 International Joint Conference on Neural Networks, IJCNN 2016., 7727303, Proceedings of the International Joint Conference on Neural Networks, vol. 2016-October, Institute of Electrical and Electronics Engineers Inc., pp. 971-976, 2016 International Joint Conference on Neural Networks, IJCNN 2016, Vancouver, Canada, 7/24/16. https://doi.org/10.1109/IJCNN.2016.7727303

On the energy benefits of spiking deep neural networks : A case study. / Han, Bing; Sengupta, Abhronil; Roy, Kaushik.

2016 International Joint Conference on Neural Networks, IJCNN 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 971-976 7727303 (Proceedings of the International Joint Conference on Neural Networks; Vol. 2016-October).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - On the energy benefits of spiking deep neural networks

T2 - A case study

AU - Han, Bing

AU - Sengupta, Abhronil

AU - Roy, Kaushik

PY - 2016/10/31

Y1 - 2016/10/31

N2 - Deep learning neural networks have achieved success in a large number of visual processing tasks and are currently utilized for many real-world applications like image search and speech recognition among others. However, in spite of achieving high accuracy in such classification problems, they involve significant computational resources. Over the past few years, artificial neural network models have evolved into the biologically realistic and event-driven spiking neural networks. Recent research efforts have been directed at developing mechanisms to convert traditional deep artificial nets to spiking nets where the neurons communicate by means of spikes. However, there have been limited studies providing insights on the specific power, area and energy benefits offered by deep spiking neural nets in comparison to their non-spiking counterparts. In this paper, we perform a case study for a hardware implementation of a spiking/non-spiking deep net on the MNIST dataset and clearly outline the design prospects involved in implementing neural computing platforms in the spiking mode of operation.

AB - Deep learning neural networks have achieved success in a large number of visual processing tasks and are currently utilized for many real-world applications like image search and speech recognition among others. However, in spite of achieving high accuracy in such classification problems, they involve significant computational resources. Over the past few years, artificial neural network models have evolved into the biologically realistic and event-driven spiking neural networks. Recent research efforts have been directed at developing mechanisms to convert traditional deep artificial nets to spiking nets where the neurons communicate by means of spikes. However, there have been limited studies providing insights on the specific power, area and energy benefits offered by deep spiking neural nets in comparison to their non-spiking counterparts. In this paper, we perform a case study for a hardware implementation of a spiking/non-spiking deep net on the MNIST dataset and clearly outline the design prospects involved in implementing neural computing platforms in the spiking mode of operation.

UR - http://www.scopus.com/inward/record.url?scp=85007277316&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85007277316&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2016.7727303

DO - 10.1109/IJCNN.2016.7727303

M3 - Conference contribution

AN - SCOPUS:85007277316

T3 - Proceedings of the International Joint Conference on Neural Networks

SP - 971

EP - 976

BT - 2016 International Joint Conference on Neural Networks, IJCNN 2016

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Han B, Sengupta A, Roy K. On the energy benefits of spiking deep neural networks: A case study. In 2016 International Joint Conference on Neural Networks, IJCNN 2016. Institute of Electrical and Electronics Engineers Inc. 2016. p. 971-976. 7727303. (Proceedings of the International Joint Conference on Neural Networks). https://doi.org/10.1109/IJCNN.2016.7727303