Challenges for large-scale implementations of spiking neural networks on FPGAs

Research output: Contribution to journalArticle

125 Citations (Scopus)

Abstract

The last 50 years has witnessed considerable research in the area of neural networks resulting in a range of architectures, learning algorithms and demonstrative applications. A more recent research trend has focused on the biological plausibility of such networks as a closer abstraction to real neurons may offer improved performance in an adaptable, real-time environment. This poses considerable challenges for engineers particularly in terms of the requirement to realise a low-cost embedded solution. Programmable hardware has been widely recognised as an ideal platform for the adaptable requirements of neural networks and there has been considerable research reported in the literature. This paper aims to review this body of research to identify the key lessons learned and, in particular, to identify the remaining challenges for large-scale implementations of spiking neural networks on FPGAs.
Original languageEnglish
Pages (from-to)13-29
Number of pages17
JournalNeurocomputing
Volume71
Issue number1
DOIs
Publication statusPublished - Dec 2007

Fingerprint

Field programmable gate arrays (FPGA)
Neural networks
Research
Learning algorithms
Neurons
Learning
Hardware
Engineers
Costs and Cost Analysis
Costs

Keywords

  • Field programmable gate arrays (FPGAs)
  • Hardware implementation
  • Spiking neural network (SNN)

Cite this

Challenges for large-scale implementations of spiking neural networks on FPGAs. / Ghani, Arfan.

In: Neurocomputing, Vol. 71, No. 1, 12.2007, p. 13-29.

Research output: Contribution to journalArticle

@article{3df5fccfef5e4e9284fc30d8667aac7b,
title = "Challenges for large-scale implementations of spiking neural networks on FPGAs",
abstract = "The last 50 years has witnessed considerable research in the area of neural networks resulting in a range of architectures, learning algorithms and demonstrative applications. A more recent research trend has focused on the biological plausibility of such networks as a closer abstraction to real neurons may offer improved performance in an adaptable, real-time environment. This poses considerable challenges for engineers particularly in terms of the requirement to realise a low-cost embedded solution. Programmable hardware has been widely recognised as an ideal platform for the adaptable requirements of neural networks and there has been considerable research reported in the literature. This paper aims to review this body of research to identify the key lessons learned and, in particular, to identify the remaining challenges for large-scale implementations of spiking neural networks on FPGAs.",
keywords = "Field programmable gate arrays (FPGAs), Hardware implementation, Spiking neural network (SNN)",
author = "Arfan Ghani",
year = "2007",
month = "12",
doi = "10.1016/j.neucom.2006.11.029",
language = "English",
volume = "71",
pages = "13--29",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Challenges for large-scale implementations of spiking neural networks on FPGAs

AU - Ghani, Arfan

PY - 2007/12

Y1 - 2007/12

N2 - The last 50 years has witnessed considerable research in the area of neural networks resulting in a range of architectures, learning algorithms and demonstrative applications. A more recent research trend has focused on the biological plausibility of such networks as a closer abstraction to real neurons may offer improved performance in an adaptable, real-time environment. This poses considerable challenges for engineers particularly in terms of the requirement to realise a low-cost embedded solution. Programmable hardware has been widely recognised as an ideal platform for the adaptable requirements of neural networks and there has been considerable research reported in the literature. This paper aims to review this body of research to identify the key lessons learned and, in particular, to identify the remaining challenges for large-scale implementations of spiking neural networks on FPGAs.

AB - The last 50 years has witnessed considerable research in the area of neural networks resulting in a range of architectures, learning algorithms and demonstrative applications. A more recent research trend has focused on the biological plausibility of such networks as a closer abstraction to real neurons may offer improved performance in an adaptable, real-time environment. This poses considerable challenges for engineers particularly in terms of the requirement to realise a low-cost embedded solution. Programmable hardware has been widely recognised as an ideal platform for the adaptable requirements of neural networks and there has been considerable research reported in the literature. This paper aims to review this body of research to identify the key lessons learned and, in particular, to identify the remaining challenges for large-scale implementations of spiking neural networks on FPGAs.

KW - Field programmable gate arrays (FPGAs)

KW - Hardware implementation

KW - Spiking neural network (SNN)

U2 - 10.1016/j.neucom.2006.11.029

DO - 10.1016/j.neucom.2006.11.029

M3 - Article

VL - 71

SP - 13

EP - 29

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 1

ER -