Abstract

In a neuromorphic computing system, the complex CMOS neuron circuits have been the bottleneck for efficient implementation of weighted sum operation. The phenomenon of metal-insulator-transition (MIT) in strongly correlated oxides, such as NbO2, has shown the oscillation behavior in recent experiments. In this work, we propose using a MIT device to function as a compact oscillation neuron, achieving the same functionality as the CMOS neuron but occupying a much smaller area. Pt/NbOx/Pt devices are fabricated, exhibiting the threshold switching I-V hysteresis. When the NbOx device is connected with an external resistor (i.e., the synapse), the neuron membrane voltage starts a self-oscillation. We experimentally demonstrate that the oscillation frequency is proportional to the conductance of the synapse, showing its feasibility for integrating the weighted sum current. The switching speed measurement indicates that the oscillation frequency could achieve >33 MHz if parasitic capacitance can be eliminated.

Original languageEnglish (US)
Article number103503
JournalApplied Physics Letters
Volume111
Issue number10
DOIs
StatePublished - Sep 4 2017

Fingerprint

neurons
synapses
oscillations
CMOS
insulators
self oscillation
resistors
metals
capacitance
hysteresis
membranes
thresholds
oxides
electric potential

ASJC Scopus subject areas

  • Physics and Astronomy (miscellaneous)

Cite this

NbOx based oscillation neuron for neuromorphic computing. / Gao, Ligang; Chen, Pai Yu; Yu, Shimeng.

In: Applied Physics Letters, Vol. 111, No. 10, 103503, 04.09.2017.

Research output: Contribution to journalArticle

Gao, Ligang ; Chen, Pai Yu ; Yu, Shimeng. / NbOx based oscillation neuron for neuromorphic computing. In: Applied Physics Letters. 2017 ; Vol. 111, No. 10.
@article{528de23a0ef74a1790ee260645348567,
title = "NbOx based oscillation neuron for neuromorphic computing",
abstract = "In a neuromorphic computing system, the complex CMOS neuron circuits have been the bottleneck for efficient implementation of weighted sum operation. The phenomenon of metal-insulator-transition (MIT) in strongly correlated oxides, such as NbO2, has shown the oscillation behavior in recent experiments. In this work, we propose using a MIT device to function as a compact oscillation neuron, achieving the same functionality as the CMOS neuron but occupying a much smaller area. Pt/NbOx/Pt devices are fabricated, exhibiting the threshold switching I-V hysteresis. When the NbOx device is connected with an external resistor (i.e., the synapse), the neuron membrane voltage starts a self-oscillation. We experimentally demonstrate that the oscillation frequency is proportional to the conductance of the synapse, showing its feasibility for integrating the weighted sum current. The switching speed measurement indicates that the oscillation frequency could achieve >33 MHz if parasitic capacitance can be eliminated.",
author = "Ligang Gao and Chen, {Pai Yu} and Shimeng Yu",
year = "2017",
month = "9",
day = "4",
doi = "10.1063/1.4991917",
language = "English (US)",
volume = "111",
journal = "Applied Physics Letters",
issn = "0003-6951",
publisher = "American Institute of Physics Publising LLC",
number = "10",

}

TY - JOUR

T1 - NbOx based oscillation neuron for neuromorphic computing

AU - Gao, Ligang

AU - Chen, Pai Yu

AU - Yu, Shimeng

PY - 2017/9/4

Y1 - 2017/9/4

N2 - In a neuromorphic computing system, the complex CMOS neuron circuits have been the bottleneck for efficient implementation of weighted sum operation. The phenomenon of metal-insulator-transition (MIT) in strongly correlated oxides, such as NbO2, has shown the oscillation behavior in recent experiments. In this work, we propose using a MIT device to function as a compact oscillation neuron, achieving the same functionality as the CMOS neuron but occupying a much smaller area. Pt/NbOx/Pt devices are fabricated, exhibiting the threshold switching I-V hysteresis. When the NbOx device is connected with an external resistor (i.e., the synapse), the neuron membrane voltage starts a self-oscillation. We experimentally demonstrate that the oscillation frequency is proportional to the conductance of the synapse, showing its feasibility for integrating the weighted sum current. The switching speed measurement indicates that the oscillation frequency could achieve >33 MHz if parasitic capacitance can be eliminated.

AB - In a neuromorphic computing system, the complex CMOS neuron circuits have been the bottleneck for efficient implementation of weighted sum operation. The phenomenon of metal-insulator-transition (MIT) in strongly correlated oxides, such as NbO2, has shown the oscillation behavior in recent experiments. In this work, we propose using a MIT device to function as a compact oscillation neuron, achieving the same functionality as the CMOS neuron but occupying a much smaller area. Pt/NbOx/Pt devices are fabricated, exhibiting the threshold switching I-V hysteresis. When the NbOx device is connected with an external resistor (i.e., the synapse), the neuron membrane voltage starts a self-oscillation. We experimentally demonstrate that the oscillation frequency is proportional to the conductance of the synapse, showing its feasibility for integrating the weighted sum current. The switching speed measurement indicates that the oscillation frequency could achieve >33 MHz if parasitic capacitance can be eliminated.

UR - http://www.scopus.com/inward/record.url?scp=85029300618&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029300618&partnerID=8YFLogxK

U2 - 10.1063/1.4991917

DO - 10.1063/1.4991917

M3 - Article

AN - SCOPUS:85029300618

VL - 111

JO - Applied Physics Letters

JF - Applied Physics Letters

SN - 0003-6951

IS - 10

M1 - 103503

ER -