An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model

The scale of modern neural networks is growing rapidly, with direct hardware implementations providing significant speed and energy improvements over their software counterparts. However, these hardware implementations frequently assume global connectivity between neurons and thus suffer from commun...

Full description

Bibliographic Details
Main Authors: Jonathan Graham-Harper-Cater, Benjamin Metcalfe, Peter Wilson
Format: Article
Language:English
Published: MDPI AG 2018-08-01
Series:Computers
Subjects:
Online Access:http://www.mdpi.com/2073-431X/7/3/43
id doaj-1b64e35fd65249eba332b2d2334a46e4
record_format Article
spelling doaj-1b64e35fd65249eba332b2d2334a46e42020-11-25T02:28:21ZengMDPI AGComputers2073-431X2018-08-01734310.3390/computers7030043computers7030043An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive ModelJonathan Graham-Harper-Cater0Benjamin Metcalfe1Peter Wilson2Department of Electronic and Electrical Engineering, University of Bath, Bath BA2 7AY, UKDepartment of Electronic and Electrical Engineering, University of Bath, Bath BA2 7AY, UKDepartment of Electronic and Electrical Engineering, University of Bath, Bath BA2 7AY, UKThe scale of modern neural networks is growing rapidly, with direct hardware implementations providing significant speed and energy improvements over their software counterparts. However, these hardware implementations frequently assume global connectivity between neurons and thus suffer from communication bottlenecks. Such issues are not found in biological neural networks. It should therefore be possible to develop new architectures to reduce the dependence on global communications by considering the connectivity of biological networks. This paper introduces two reconfigurable locally-connected architectures for implementing biologically inspired neural networks in real time. Both proposed architectures are validated using the segmented locomotive model of the C. elegans, performing a demonstration of forwards, backwards serpentine motion and coiling behaviours. Local connectivity is discovered to offer up to a 17.5× speed improvement over hybrid systems that use combinations of local and global infrastructure. Furthermore, the concept of locality of connections is considered in more detail, highlighting the importance of dimensionality when designing neuromorphic architectures. Convolutional Neural Networks are shown to map poorly to locally connected architectures despite their apparent local structure, and both the locality and dimensionality of new neural processing systems is demonstrated as a critical component for matching the function and efficiency seen in biological networks.http://www.mdpi.com/2073-431X/7/3/43reconfigurablearchitectureneural-networkneuromorphicFPGAC. elegans
collection DOAJ
language English
format Article
sources DOAJ
author Jonathan Graham-Harper-Cater
Benjamin Metcalfe
Peter Wilson
spellingShingle Jonathan Graham-Harper-Cater
Benjamin Metcalfe
Peter Wilson
An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
Computers
reconfigurable
architecture
neural-network
neuromorphic
FPGA
C. elegans
author_facet Jonathan Graham-Harper-Cater
Benjamin Metcalfe
Peter Wilson
author_sort Jonathan Graham-Harper-Cater
title An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
title_short An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
title_full An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
title_fullStr An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
title_full_unstemmed An Analytical Comparison of Locally-Connected Reconfigurable Neural Network Architectures Using a C. elegans Locomotive Model
title_sort analytical comparison of locally-connected reconfigurable neural network architectures using a c. elegans locomotive model
publisher MDPI AG
series Computers
issn 2073-431X
publishDate 2018-08-01
description The scale of modern neural networks is growing rapidly, with direct hardware implementations providing significant speed and energy improvements over their software counterparts. However, these hardware implementations frequently assume global connectivity between neurons and thus suffer from communication bottlenecks. Such issues are not found in biological neural networks. It should therefore be possible to develop new architectures to reduce the dependence on global communications by considering the connectivity of biological networks. This paper introduces two reconfigurable locally-connected architectures for implementing biologically inspired neural networks in real time. Both proposed architectures are validated using the segmented locomotive model of the C. elegans, performing a demonstration of forwards, backwards serpentine motion and coiling behaviours. Local connectivity is discovered to offer up to a 17.5× speed improvement over hybrid systems that use combinations of local and global infrastructure. Furthermore, the concept of locality of connections is considered in more detail, highlighting the importance of dimensionality when designing neuromorphic architectures. Convolutional Neural Networks are shown to map poorly to locally connected architectures despite their apparent local structure, and both the locality and dimensionality of new neural processing systems is demonstrated as a critical component for matching the function and efficiency seen in biological networks.
topic reconfigurable
architecture
neural-network
neuromorphic
FPGA
C. elegans
url http://www.mdpi.com/2073-431X/7/3/43
work_keys_str_mv AT jonathangrahamharpercater ananalyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
AT benjaminmetcalfe ananalyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
AT peterwilson ananalyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
AT jonathangrahamharpercater analyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
AT benjaminmetcalfe analyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
AT peterwilson analyticalcomparisonoflocallyconnectedreconfigurableneuralnetworkarchitecturesusingaceleganslocomotivemodel
_version_ 1724838693317902336