Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems

Bin Gao, Yingjie Bi, Hong Yu Chen, Rui Liu, Peng Huang, Bing Chen, Lifeng Liu, Xiaoyan Liu, Shimeng Yu, H. S Philip Wong, Jinfeng Kang

Research output: Contribution to journalArticle

88 Citations (Scopus)

Abstract

Neuromorphic computing is an attractive computation paradigm that complements the von Neumann architecture. The salient features of neuromorphic computing are massive parallelism, adaptivity to the complex input information, and tolerance to errors. As one of the most crucial components in a neuromorphic system, the electronic synapse requires high device integration density and low-energy consumption. Oxide-based resistive switching devices have been shown to be a promising candidate to realize the functions of the synapse. However, the intrinsic variation increases significantly with the reduced spike energy due to the reduced number of oxygen vacancies in the conductive filament region. The large resistance variation may degrade the accuracy of neuromorphic computation. In this work, we develop an oxide-based electronic synapse to suppress the degradation caused by the intrinsic resistance variation. The synapse utilizes a three-dimensional vertical structure including several parallel oxide-based resistive switching devices on the same nanopillar. The fabricated three-dimensional electronic synapse exhibits the potential for low fabrication cost, high integration density, and excellent performances, such as low training energy per spike, gradual resistance transition under identical pulse training scheme, and good repeatability. A pattern recognition computation is simulated based on a well-known neuromorphic visual system to quantify the feasibility of the three-dimensional vertical structured synapse for the application of neuromorphic computation systems. The simulation results show significantly improved recognition accuracy from 65 to 90% after introducing the three-dimensional synapses.

Original languageEnglish (US)
Pages (from-to)6998-7004
Number of pages7
JournalACS Nano
Volume8
Issue number7
DOIs
StatePublished - Jul 22 2014

Fingerprint

synapses
Oxides
oxides
electronics
energy
spikes
Oxygen vacancies
education
Pattern recognition
Energy utilization
Fabrication
Degradation
energy consumption
pattern recognition
complement
filaments
Costs
degradation
costs
fabrication

Keywords

  • 3D integration
  • memory
  • metal oxide
  • neuromorphic computation
  • resistive switching
  • synapse
  • synaptic device

ASJC Scopus subject areas

  • Engineering(all)
  • Materials Science(all)
  • Physics and Astronomy(all)

Cite this

Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems. / Gao, Bin; Bi, Yingjie; Chen, Hong Yu; Liu, Rui; Huang, Peng; Chen, Bing; Liu, Lifeng; Liu, Xiaoyan; Yu, Shimeng; Wong, H. S Philip; Kang, Jinfeng.

In: ACS Nano, Vol. 8, No. 7, 22.07.2014, p. 6998-7004.

Research output: Contribution to journalArticle

Gao, B, Bi, Y, Chen, HY, Liu, R, Huang, P, Chen, B, Liu, L, Liu, X, Yu, S, Wong, HSP & Kang, J 2014, 'Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems', ACS Nano, vol. 8, no. 7, pp. 6998-7004. https://doi.org/10.1021/nn501824r
Gao, Bin ; Bi, Yingjie ; Chen, Hong Yu ; Liu, Rui ; Huang, Peng ; Chen, Bing ; Liu, Lifeng ; Liu, Xiaoyan ; Yu, Shimeng ; Wong, H. S Philip ; Kang, Jinfeng. / Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems. In: ACS Nano. 2014 ; Vol. 8, No. 7. pp. 6998-7004.
@article{2831cfd526624cfe90c01833e5350ce4,
title = "Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems",
abstract = "Neuromorphic computing is an attractive computation paradigm that complements the von Neumann architecture. The salient features of neuromorphic computing are massive parallelism, adaptivity to the complex input information, and tolerance to errors. As one of the most crucial components in a neuromorphic system, the electronic synapse requires high device integration density and low-energy consumption. Oxide-based resistive switching devices have been shown to be a promising candidate to realize the functions of the synapse. However, the intrinsic variation increases significantly with the reduced spike energy due to the reduced number of oxygen vacancies in the conductive filament region. The large resistance variation may degrade the accuracy of neuromorphic computation. In this work, we develop an oxide-based electronic synapse to suppress the degradation caused by the intrinsic resistance variation. The synapse utilizes a three-dimensional vertical structure including several parallel oxide-based resistive switching devices on the same nanopillar. The fabricated three-dimensional electronic synapse exhibits the potential for low fabrication cost, high integration density, and excellent performances, such as low training energy per spike, gradual resistance transition under identical pulse training scheme, and good repeatability. A pattern recognition computation is simulated based on a well-known neuromorphic visual system to quantify the feasibility of the three-dimensional vertical structured synapse for the application of neuromorphic computation systems. The simulation results show significantly improved recognition accuracy from 65 to 90{\%} after introducing the three-dimensional synapses.",
keywords = "3D integration, memory, metal oxide, neuromorphic computation, resistive switching, synapse, synaptic device",
author = "Bin Gao and Yingjie Bi and Chen, {Hong Yu} and Rui Liu and Peng Huang and Bing Chen and Lifeng Liu and Xiaoyan Liu and Shimeng Yu and Wong, {H. S Philip} and Jinfeng Kang",
year = "2014",
month = "7",
day = "22",
doi = "10.1021/nn501824r",
language = "English (US)",
volume = "8",
pages = "6998--7004",
journal = "ACS Nano",
issn = "1936-0851",
publisher = "American Chemical Society",
number = "7",

}

TY - JOUR

T1 - Ultra-low-energy three-dimensional oxide-based electronic synapses for implementation of robust high-accuracy neuromorphic computation systems

AU - Gao, Bin

AU - Bi, Yingjie

AU - Chen, Hong Yu

AU - Liu, Rui

AU - Huang, Peng

AU - Chen, Bing

AU - Liu, Lifeng

AU - Liu, Xiaoyan

AU - Yu, Shimeng

AU - Wong, H. S Philip

AU - Kang, Jinfeng

PY - 2014/7/22

Y1 - 2014/7/22

N2 - Neuromorphic computing is an attractive computation paradigm that complements the von Neumann architecture. The salient features of neuromorphic computing are massive parallelism, adaptivity to the complex input information, and tolerance to errors. As one of the most crucial components in a neuromorphic system, the electronic synapse requires high device integration density and low-energy consumption. Oxide-based resistive switching devices have been shown to be a promising candidate to realize the functions of the synapse. However, the intrinsic variation increases significantly with the reduced spike energy due to the reduced number of oxygen vacancies in the conductive filament region. The large resistance variation may degrade the accuracy of neuromorphic computation. In this work, we develop an oxide-based electronic synapse to suppress the degradation caused by the intrinsic resistance variation. The synapse utilizes a three-dimensional vertical structure including several parallel oxide-based resistive switching devices on the same nanopillar. The fabricated three-dimensional electronic synapse exhibits the potential for low fabrication cost, high integration density, and excellent performances, such as low training energy per spike, gradual resistance transition under identical pulse training scheme, and good repeatability. A pattern recognition computation is simulated based on a well-known neuromorphic visual system to quantify the feasibility of the three-dimensional vertical structured synapse for the application of neuromorphic computation systems. The simulation results show significantly improved recognition accuracy from 65 to 90% after introducing the three-dimensional synapses.

AB - Neuromorphic computing is an attractive computation paradigm that complements the von Neumann architecture. The salient features of neuromorphic computing are massive parallelism, adaptivity to the complex input information, and tolerance to errors. As one of the most crucial components in a neuromorphic system, the electronic synapse requires high device integration density and low-energy consumption. Oxide-based resistive switching devices have been shown to be a promising candidate to realize the functions of the synapse. However, the intrinsic variation increases significantly with the reduced spike energy due to the reduced number of oxygen vacancies in the conductive filament region. The large resistance variation may degrade the accuracy of neuromorphic computation. In this work, we develop an oxide-based electronic synapse to suppress the degradation caused by the intrinsic resistance variation. The synapse utilizes a three-dimensional vertical structure including several parallel oxide-based resistive switching devices on the same nanopillar. The fabricated three-dimensional electronic synapse exhibits the potential for low fabrication cost, high integration density, and excellent performances, such as low training energy per spike, gradual resistance transition under identical pulse training scheme, and good repeatability. A pattern recognition computation is simulated based on a well-known neuromorphic visual system to quantify the feasibility of the three-dimensional vertical structured synapse for the application of neuromorphic computation systems. The simulation results show significantly improved recognition accuracy from 65 to 90% after introducing the three-dimensional synapses.

KW - 3D integration

KW - memory

KW - metal oxide

KW - neuromorphic computation

KW - resistive switching

KW - synapse

KW - synaptic device

UR - http://www.scopus.com/inward/record.url?scp=84904757475&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84904757475&partnerID=8YFLogxK

U2 - 10.1021/nn501824r

DO - 10.1021/nn501824r

M3 - Article

C2 - 24884237

AN - SCOPUS:84904757475

VL - 8

SP - 6998

EP - 7004

JO - ACS Nano

JF - ACS Nano

SN - 1936-0851

IS - 7

ER -