Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems

Amr M.S. Tosson, Shimeng Yu, Mohab Anis, Lan Wei

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

2 Citations (Scopus)

Abstract

With the speed and power bottleneck in the conventional Von Neumann architecture, the interest in the neuromorphic systems has greatly increased in recent years. To create a highly dense communication network between the preand post-neurons, RRAM devices are used as synapses in the neuromorphic systems due to many advantages including their small sizes and low-power operations. However, due to RRAM reliability issues, in particular soft-errors, the performance of the RRAM-based neuromorphic systems are significantly degraded. In this article, we propose a novel framework for detecting and resolving the degradation in the system performance due to the RRAM reliability soft-errors. The read and write circuits modifications to implement the framework, and their impact on the delay and energy consumption of the neuromorphic system are also discussed in this article. Using a combination of BRIAN and SPICE simulations, we demonstrate that the proposed framework can restore the accuracy of the example RRAM-based neuromorphic system from 43% back to its target value of 91.6% with a minimal impact on the read (< 0.1% and 1.1% increase in the delay and energy respectively) and write (0% and < 0.1% increase in the delay and energy respectively) operations.

Original languageEnglish (US)
Title of host publicationGLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017
PublisherAssociation for Computing Machinery
Pages53-58
Number of pages6
VolumePart F127756
ISBN (Electronic)9781450349727
DOIs
StatePublished - May 10 2017
Event27th Great Lakes Symposium on VLSI, GLSVLSI 2017 - Banff, Canada
Duration: May 10 2017May 12 2017

Other

Other27th Great Lakes Symposium on VLSI, GLSVLSI 2017
CountryCanada
CityBanff
Period5/10/175/12/17

Fingerprint

SPICE
Neurons
Telecommunication networks
Energy utilization
RRAM
Degradation
Networks (circuits)

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Tosson, A. M. S., Yu, S., Anis, M., & Wei, L. (2017). Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems. In GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017 (Vol. Part F127756, pp. 53-58). Association for Computing Machinery. https://doi.org/10.1145/3060403.3060431

Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems. / Tosson, Amr M.S.; Yu, Shimeng; Anis, Mohab; Wei, Lan.

GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017. Vol. Part F127756 Association for Computing Machinery, 2017. p. 53-58.

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

Tosson, AMS, Yu, S, Anis, M & Wei, L 2017, Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems. in GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017. vol. Part F127756, Association for Computing Machinery, pp. 53-58, 27th Great Lakes Symposium on VLSI, GLSVLSI 2017, Banff, Canada, 5/10/17. https://doi.org/10.1145/3060403.3060431
Tosson AMS, Yu S, Anis M, Wei L. Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems. In GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017. Vol. Part F127756. Association for Computing Machinery. 2017. p. 53-58 https://doi.org/10.1145/3060403.3060431
Tosson, Amr M.S. ; Yu, Shimeng ; Anis, Mohab ; Wei, Lan. / Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems. GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017. Vol. Part F127756 Association for Computing Machinery, 2017. pp. 53-58
@inproceedings{768248ccd67a4bb78162f0ffafb942ab,
title = "Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems",
abstract = "With the speed and power bottleneck in the conventional Von Neumann architecture, the interest in the neuromorphic systems has greatly increased in recent years. To create a highly dense communication network between the preand post-neurons, RRAM devices are used as synapses in the neuromorphic systems due to many advantages including their small sizes and low-power operations. However, due to RRAM reliability issues, in particular soft-errors, the performance of the RRAM-based neuromorphic systems are significantly degraded. In this article, we propose a novel framework for detecting and resolving the degradation in the system performance due to the RRAM reliability soft-errors. The read and write circuits modifications to implement the framework, and their impact on the delay and energy consumption of the neuromorphic system are also discussed in this article. Using a combination of BRIAN and SPICE simulations, we demonstrate that the proposed framework can restore the accuracy of the example RRAM-based neuromorphic system from 43{\%} back to its target value of 91.6{\%} with a minimal impact on the read (< 0.1{\%} and 1.1{\%} increase in the delay and energy respectively) and write (0{\%} and < 0.1{\%} increase in the delay and energy respectively) operations.",
author = "Tosson, {Amr M.S.} and Shimeng Yu and Mohab Anis and Lan Wei",
year = "2017",
month = "5",
day = "10",
doi = "10.1145/3060403.3060431",
language = "English (US)",
volume = "Part F127756",
pages = "53--58",
booktitle = "GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - Mitigating the effect of reliability soft-errors of RRAM devices on the performance of RRAM-based neuromorphic systems

AU - Tosson, Amr M.S.

AU - Yu, Shimeng

AU - Anis, Mohab

AU - Wei, Lan

PY - 2017/5/10

Y1 - 2017/5/10

N2 - With the speed and power bottleneck in the conventional Von Neumann architecture, the interest in the neuromorphic systems has greatly increased in recent years. To create a highly dense communication network between the preand post-neurons, RRAM devices are used as synapses in the neuromorphic systems due to many advantages including their small sizes and low-power operations. However, due to RRAM reliability issues, in particular soft-errors, the performance of the RRAM-based neuromorphic systems are significantly degraded. In this article, we propose a novel framework for detecting and resolving the degradation in the system performance due to the RRAM reliability soft-errors. The read and write circuits modifications to implement the framework, and their impact on the delay and energy consumption of the neuromorphic system are also discussed in this article. Using a combination of BRIAN and SPICE simulations, we demonstrate that the proposed framework can restore the accuracy of the example RRAM-based neuromorphic system from 43% back to its target value of 91.6% with a minimal impact on the read (< 0.1% and 1.1% increase in the delay and energy respectively) and write (0% and < 0.1% increase in the delay and energy respectively) operations.

AB - With the speed and power bottleneck in the conventional Von Neumann architecture, the interest in the neuromorphic systems has greatly increased in recent years. To create a highly dense communication network between the preand post-neurons, RRAM devices are used as synapses in the neuromorphic systems due to many advantages including their small sizes and low-power operations. However, due to RRAM reliability issues, in particular soft-errors, the performance of the RRAM-based neuromorphic systems are significantly degraded. In this article, we propose a novel framework for detecting and resolving the degradation in the system performance due to the RRAM reliability soft-errors. The read and write circuits modifications to implement the framework, and their impact on the delay and energy consumption of the neuromorphic system are also discussed in this article. Using a combination of BRIAN and SPICE simulations, we demonstrate that the proposed framework can restore the accuracy of the example RRAM-based neuromorphic system from 43% back to its target value of 91.6% with a minimal impact on the read (< 0.1% and 1.1% increase in the delay and energy respectively) and write (0% and < 0.1% increase in the delay and energy respectively) operations.

UR - http://www.scopus.com/inward/record.url?scp=85021194696&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021194696&partnerID=8YFLogxK

U2 - 10.1145/3060403.3060431

DO - 10.1145/3060403.3060431

M3 - Conference contribution

AN - SCOPUS:85021194696

VL - Part F127756

SP - 53

EP - 58

BT - GLSVLSI 2017 - Proceedings of the Great Lakes Symposium on VLSI 2017

PB - Association for Computing Machinery

ER -