Approximate Acceleration for CNN-based Applications on IoT Edge Devices

Jorge Castro-Godinez; Deykel Hernandez-Araya; Muhammad Shafique; Jorg Henkel

doi:10.1109/LASCAS45839.2020.9069040

Approximate Acceleration for CNN-based Applications on IoT Edge Devices

Jorge Castro-Godinez
, Deykel Hernandez-Araya
, Muhammad Shafique
, Jorg Henkel

Escuela de Ingeniería Electrónica

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

16 Scopus citations

Abstract

Machine learning based sub-systems are increasingly becoming part of IoT edge devices, thereby requiring resource-efficient architectures and implementations, especially when subjected to battery-constrained scenarios. The non-exact nature of Convolutional Neural Networks (CNNs) opens the possibility to use approximate computations to reduce their required runtime and energy consumption on resource-constrained IoT edge devices without significantly compromising their classification output. In this paper, we propose a resilience exploration method and a novel approximate accelerator to speed up the execution of the convolutional layer, which is the most time consuming component of CNNs, for IoT edge devices. Trained CNNs with Caffe framework are executed on a System-on-Chip with reconfigurable hardware available, where the approximate accelerator is deployed. CNN applications developed with Caffe can take advantage of our proposed approximate acceleration to execute them on IoT edge devices.

Original language	English
Title of host publication	2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728134277
DOIs	https://doi.org/10.1109/LASCAS45839.2020.9069040
State	Published - Feb 2020
Event	11th IEEE Latin American Symposium on Circuits and Systems, LASCAS 2020 - San Jose, Costa Rica Duration: 25 Feb 2020 → 28 Feb 2020

Publication series

Name	2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020

Conference

Conference	11th IEEE Latin American Symposium on Circuits and Systems, LASCAS 2020
Country/Territory	Costa Rica
City	San Jose
Period	25/02/20 → 28/02/20

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

Approximate computing
accelerator architectures
convolutional neural networks
edge computing

Access to Document

10.1109/LASCAS45839.2020.9069040

Cite this

Castro-Godinez, J., Hernandez-Araya, D., Shafique, M., & Henkel, J. (2020). Approximate Acceleration for CNN-based Applications on IoT Edge Devices. In 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020 Article 9069040 (2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/LASCAS45839.2020.9069040

Castro-Godinez, Jorge ; Hernandez-Araya, Deykel ; Shafique, Muhammad et al. / Approximate Acceleration for CNN-based Applications on IoT Edge Devices. 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020. Institute of Electrical and Electronics Engineers Inc., 2020. (2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020).

@inproceedings{830067b223b042eab4f6ae8d698d7fb4,

title = "Approximate Acceleration for CNN-based Applications on IoT Edge Devices",

abstract = "Machine learning based sub-systems are increasingly becoming part of IoT edge devices, thereby requiring resource-efficient architectures and implementations, especially when subjected to battery-constrained scenarios. The non-exact nature of Convolutional Neural Networks (CNNs) opens the possibility to use approximate computations to reduce their required runtime and energy consumption on resource-constrained IoT edge devices without significantly compromising their classification output. In this paper, we propose a resilience exploration method and a novel approximate accelerator to speed up the execution of the convolutional layer, which is the most time consuming component of CNNs, for IoT edge devices. Trained CNNs with Caffe framework are executed on a System-on-Chip with reconfigurable hardware available, where the approximate accelerator is deployed. CNN applications developed with Caffe can take advantage of our proposed approximate acceleration to execute them on IoT edge devices.",

keywords = "Approximate computing, accelerator architectures, convolutional neural networks, edge computing",

author = "Jorge Castro-Godinez and Deykel Hernandez-Araya and Muhammad Shafique and Jorg Henkel",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 11th IEEE Latin American Symposium on Circuits and Systems, LASCAS 2020 ; Conference date: 25-02-2020 Through 28-02-2020",

year = "2020",

month = feb,

doi = "10.1109/LASCAS45839.2020.9069040",

language = "Ingl{\'e}s",

series = "2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020",

}

Castro-Godinez, J, Hernandez-Araya, D, Shafique, M & Henkel, J 2020, Approximate Acceleration for CNN-based Applications on IoT Edge Devices. in 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020., 9069040, 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020, Institute of Electrical and Electronics Engineers Inc., 11th IEEE Latin American Symposium on Circuits and Systems, LASCAS 2020, San Jose, Costa Rica, 25/02/20. https://doi.org/10.1109/LASCAS45839.2020.9069040

Approximate Acceleration for CNN-based Applications on IoT Edge Devices. / Castro-Godinez, Jorge; Hernandez-Araya, Deykel; Shafique, Muhammad et al.
2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 9069040 (2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020).

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

TY - GEN

T1 - Approximate Acceleration for CNN-based Applications on IoT Edge Devices

AU - Castro-Godinez, Jorge

AU - Hernandez-Araya, Deykel

AU - Shafique, Muhammad

AU - Henkel, Jorg

PY - 2020/2

Y1 - 2020/2

N2 - Machine learning based sub-systems are increasingly becoming part of IoT edge devices, thereby requiring resource-efficient architectures and implementations, especially when subjected to battery-constrained scenarios. The non-exact nature of Convolutional Neural Networks (CNNs) opens the possibility to use approximate computations to reduce their required runtime and energy consumption on resource-constrained IoT edge devices without significantly compromising their classification output. In this paper, we propose a resilience exploration method and a novel approximate accelerator to speed up the execution of the convolutional layer, which is the most time consuming component of CNNs, for IoT edge devices. Trained CNNs with Caffe framework are executed on a System-on-Chip with reconfigurable hardware available, where the approximate accelerator is deployed. CNN applications developed with Caffe can take advantage of our proposed approximate acceleration to execute them on IoT edge devices.

AB - Machine learning based sub-systems are increasingly becoming part of IoT edge devices, thereby requiring resource-efficient architectures and implementations, especially when subjected to battery-constrained scenarios. The non-exact nature of Convolutional Neural Networks (CNNs) opens the possibility to use approximate computations to reduce their required runtime and energy consumption on resource-constrained IoT edge devices without significantly compromising their classification output. In this paper, we propose a resilience exploration method and a novel approximate accelerator to speed up the execution of the convolutional layer, which is the most time consuming component of CNNs, for IoT edge devices. Trained CNNs with Caffe framework are executed on a System-on-Chip with reconfigurable hardware available, where the approximate accelerator is deployed. CNN applications developed with Caffe can take advantage of our proposed approximate acceleration to execute them on IoT edge devices.

KW - Approximate computing

KW - accelerator architectures

KW - convolutional neural networks

KW - edge computing

UR - https://www.scopus.com/pages/publications/85084325560

U2 - 10.1109/LASCAS45839.2020.9069040

DO - 10.1109/LASCAS45839.2020.9069040

M3 - Contribución a la conferencia

AN - SCOPUS:85084325560

T3 - 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020

BT - 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 11th IEEE Latin American Symposium on Circuits and Systems, LASCAS 2020

Y2 - 25 February 2020 through 28 February 2020

ER -

Castro-Godinez J, Hernandez-Araya D, Shafique M, Henkel J. Approximate Acceleration for CNN-based Applications on IoT Edge Devices. In 2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020. Institute of Electrical and Electronics Engineers Inc. 2020. 9069040. (2020 IEEE 11th Latin American Symposium on Circuits and Systems, LASCAS 2020). doi: 10.1109/LASCAS45839.2020.9069040

Approximate Acceleration for CNN-based Applications on IoT Edge Devices

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this