Accelerating Spiking Neural Networks using Memristive Crossbar Arrays

Thomas Bohnstingl*, Angeliki Pantazi, Evangelos Eleftheriou

*Korrespondierende/r Autor/in für diese Arbeit

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem Konferenzband

Abstract

Biologically-inspiredspikingneuralnetworks(SNNs) hold great promise to perform demanding tasks in an energy and area-efficient manner. Memristive devices organized in a crossbar array can be used to accelerate operations of artificial neural networks (ANNs) while circumventing limitations of traditional computing paradigms. Recent advances have led to the development of neuromorphic accelerators that employ phase-change memory (PCM) devices. We propose an approach to fully unravel the potential of such systems for SNNs by integrating entire layers, including synaptic weights as well as neuronal states, into crossbar arrays. However,the key challenges of such realizations originate from the intrinsic imperfections of the PCM devices that limit their effective precision. Thus, we investigated the impact of these limitations on the performance of SNNs and demonstrate that the synaptic weight and neuronal state realization using 4-bitprecision provides a robust network performance. Moreover, we evaluated the scheme for a multi-layer SNN realized using an experimentally verified model of the PCM devices and achieved performance that is comparable to a floating-point 32-bit model.
Originalspracheenglisch
TitelICECS 2020 - 27th IEEE International Conference on Electronics, Circuits and Systems, Proceedings
ISBN (elektronisch)9781728160443
DOIs
PublikationsstatusVeröffentlicht - 23 Nov 2020
Veranstaltung27th IEEE International Conference on Electronics, Circuits and Systems: ICECS 2020 - Virtual, Glasgow, Großbritannien / Vereinigtes Königreich
Dauer: 23 Nov 202025 Nov 2020

Konferenz

Konferenz27th IEEE International Conference on Electronics, Circuits and Systems
LandGroßbritannien / Vereinigtes Königreich
OrtVirtual, Glasgow
Zeitraum23/11/2025/11/20

ASJC Scopus subject areas

  • !!Electrical and Electronic Engineering

Fingerprint Untersuchen Sie die Forschungsthemen von „Accelerating Spiking Neural Networks using Memristive Crossbar Arrays“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren