A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware

Artificial neural networks (ANNs), like convolutional neural networks (CNNs), have achieved the state-of-the-art results for many machine learning tasks. However, inference with large-scale full-precision CNNs must cause substantial energy consumption and memory occupation, which seriously hinders t...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Chenglong Zou, Xiaoxin Cui, Yisong Kuang, Kefei Liu, Yuan Wang, Xinan Wang, Ru Huang
Formato: article
Lenguaje:EN
Publicado: Frontiers Media S.A. 2021
Materias:
Acceso en línea:https://doaj.org/article/6e4eedc885474c579ffec00f6a476c8a
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:6e4eedc885474c579ffec00f6a476c8a
record_format dspace
spelling oai:doaj.org-article:6e4eedc885474c579ffec00f6a476c8a2021-11-16T07:39:16ZA Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware1662-453X10.3389/fnins.2021.694170https://doaj.org/article/6e4eedc885474c579ffec00f6a476c8a2021-11-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fnins.2021.694170/fullhttps://doaj.org/toc/1662-453XArtificial neural networks (ANNs), like convolutional neural networks (CNNs), have achieved the state-of-the-art results for many machine learning tasks. However, inference with large-scale full-precision CNNs must cause substantial energy consumption and memory occupation, which seriously hinders their deployment on mobile and embedded systems. Highly inspired from biological brain, spiking neural networks (SNNs) are emerging as new solutions because of natural superiority in brain-like learning and great energy efficiency with event-driven communication and computation. Nevertheless, training a deep SNN remains a main challenge and there is usually a big accuracy gap between ANNs and SNNs. In this paper, we introduce a hardware-friendly conversion algorithm called “scatter-and-gather” to convert quantized ANNs to lossless SNNs, where neurons are connected with ternary {−1,0,1} synaptic weights. Each spiking neuron is stateless and more like original McCulloch and Pitts model, because it fires at most one spike and need be reset at each time step. Furthermore, we develop an incremental mapping framework to demonstrate efficient network deployments on a reconfigurable neuromorphic chip. Experimental results show our spiking LeNet on MNIST and VGG-Net on CIFAR-10 datasetobtain 99.37% and 91.91% classification accuracy, respectively. Besides, the presented mapping algorithm manages network deployment on our neuromorphic chip with maximum resource efficiency and excellent flexibility. Our four-spike LeNet and VGG-Net on chip can achieve respective real-time inference speed of 0.38 ms/image, 3.24 ms/image, and an average power consumption of 0.28 mJ/image and 2.3 mJ/image at 0.9 V, 252 MHz, which is nearly two orders of magnitude more efficient than traditional GPUs.Chenglong ZouChenglong ZouXiaoxin CuiYisong KuangKefei LiuYuan WangXinan WangRu HuangFrontiers Media S.A.articleconvolutional neural networkspiking neural networknetwork quantizationnetwork conversionneuromorphic hardwarenetwork mappingNeurosciences. Biological psychiatry. NeuropsychiatryRC321-571ENFrontiers in Neuroscience, Vol 15 (2021)
institution DOAJ
collection DOAJ
language EN
topic convolutional neural network
spiking neural network
network quantization
network conversion
neuromorphic hardware
network mapping
Neurosciences. Biological psychiatry. Neuropsychiatry
RC321-571
spellingShingle convolutional neural network
spiking neural network
network quantization
network conversion
neuromorphic hardware
network mapping
Neurosciences. Biological psychiatry. Neuropsychiatry
RC321-571
Chenglong Zou
Chenglong Zou
Xiaoxin Cui
Yisong Kuang
Kefei Liu
Yuan Wang
Xinan Wang
Ru Huang
A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
description Artificial neural networks (ANNs), like convolutional neural networks (CNNs), have achieved the state-of-the-art results for many machine learning tasks. However, inference with large-scale full-precision CNNs must cause substantial energy consumption and memory occupation, which seriously hinders their deployment on mobile and embedded systems. Highly inspired from biological brain, spiking neural networks (SNNs) are emerging as new solutions because of natural superiority in brain-like learning and great energy efficiency with event-driven communication and computation. Nevertheless, training a deep SNN remains a main challenge and there is usually a big accuracy gap between ANNs and SNNs. In this paper, we introduce a hardware-friendly conversion algorithm called “scatter-and-gather” to convert quantized ANNs to lossless SNNs, where neurons are connected with ternary {−1,0,1} synaptic weights. Each spiking neuron is stateless and more like original McCulloch and Pitts model, because it fires at most one spike and need be reset at each time step. Furthermore, we develop an incremental mapping framework to demonstrate efficient network deployments on a reconfigurable neuromorphic chip. Experimental results show our spiking LeNet on MNIST and VGG-Net on CIFAR-10 datasetobtain 99.37% and 91.91% classification accuracy, respectively. Besides, the presented mapping algorithm manages network deployment on our neuromorphic chip with maximum resource efficiency and excellent flexibility. Our four-spike LeNet and VGG-Net on chip can achieve respective real-time inference speed of 0.38 ms/image, 3.24 ms/image, and an average power consumption of 0.28 mJ/image and 2.3 mJ/image at 0.9 V, 252 MHz, which is nearly two orders of magnitude more efficient than traditional GPUs.
format article
author Chenglong Zou
Chenglong Zou
Xiaoxin Cui
Yisong Kuang
Kefei Liu
Yuan Wang
Xinan Wang
Ru Huang
author_facet Chenglong Zou
Chenglong Zou
Xiaoxin Cui
Yisong Kuang
Kefei Liu
Yuan Wang
Xinan Wang
Ru Huang
author_sort Chenglong Zou
title A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
title_short A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
title_full A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
title_fullStr A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
title_full_unstemmed A Scatter-and-Gather Spiking Convolutional Neural Network on a Reconfigurable Neuromorphic Hardware
title_sort scatter-and-gather spiking convolutional neural network on a reconfigurable neuromorphic hardware
publisher Frontiers Media S.A.
publishDate 2021
url https://doaj.org/article/6e4eedc885474c579ffec00f6a476c8a
work_keys_str_mv AT chenglongzou ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT chenglongzou ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT xiaoxincui ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT yisongkuang ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT kefeiliu ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT yuanwang ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT xinanwang ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT ruhuang ascatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT chenglongzou scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT chenglongzou scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT xiaoxincui scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT yisongkuang scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT kefeiliu scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT yuanwang scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT xinanwang scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
AT ruhuang scatterandgatherspikingconvolutionalneuralnetworkonareconfigurableneuromorphichardware
_version_ 1718426609582604288