Proposal for an all-spin artificial neural network: Emulating neural and synaptic functionalities through domain wall motion in ferromagnets

Abhronil Sengupta, Yong Shim, Kaushik Roy

Research output: Contribution to journalArticle

48 Citations (Scopus)

Abstract

Non-Boolean computing based on emerging post-CMOS technologies can potentially pave the way for low-power neural computing platforms. However, existing work on such emerging neuromorphic architectures have either focused on solely mimicking the neuron, or the synapse functionality. While memristive devices have been proposed to emulate biological synapses, spintronic devices have proved to be efficient at performing the thresholding operation of the neuron at ultra-low currents. In this work, we propose an All-Spin Artificial Neural Network where a single spintronic device acts as the basic building block of the system. The device offers a direct mapping to synapse and neuron functionalities in the brain while inter-layer network communication is accomplished via CMOS transistors. To the best of our knowledge, this is the first demonstration of a neural architecture where a single nanoelectronic device is able to mimic both neurons and synapses. The ultra-low voltage operation of low resistance magneto-metallic neurons enables the low-voltage operation of the array of spintronic synapses, thereby leading to ultra-low power neural architectures. Device-level simulations, calibrated to experimental results, was used to drive the circuit and system level simulations of the neural network for a standard pattern recognition problem. Simulation studies indicate energy savings by ∼100× in comparison to a corresponding digital/analog CMOS neuron implementation.

Original languageEnglish (US)
Article number7470633
Pages (from-to)1152-1160
Number of pages9
JournalIEEE Transactions on Biomedical Circuits and Systems
Volume10
Issue number6
DOIs
StatePublished - Dec 2016

Fingerprint

Domain walls
Neurons
Neural networks
Magnetoelectronics
Nanoelectronics
Network layers
Electric potential
Telecommunication networks
Pattern recognition
Brain
Energy conservation
Transistors
Demonstrations
Networks (circuits)

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering
  • Electrical and Electronic Engineering

Cite this

@article{254e9d88b4a04029be678b759188b16c,
title = "Proposal for an all-spin artificial neural network: Emulating neural and synaptic functionalities through domain wall motion in ferromagnets",
abstract = "Non-Boolean computing based on emerging post-CMOS technologies can potentially pave the way for low-power neural computing platforms. However, existing work on such emerging neuromorphic architectures have either focused on solely mimicking the neuron, or the synapse functionality. While memristive devices have been proposed to emulate biological synapses, spintronic devices have proved to be efficient at performing the thresholding operation of the neuron at ultra-low currents. In this work, we propose an All-Spin Artificial Neural Network where a single spintronic device acts as the basic building block of the system. The device offers a direct mapping to synapse and neuron functionalities in the brain while inter-layer network communication is accomplished via CMOS transistors. To the best of our knowledge, this is the first demonstration of a neural architecture where a single nanoelectronic device is able to mimic both neurons and synapses. The ultra-low voltage operation of low resistance magneto-metallic neurons enables the low-voltage operation of the array of spintronic synapses, thereby leading to ultra-low power neural architectures. Device-level simulations, calibrated to experimental results, was used to drive the circuit and system level simulations of the neural network for a standard pattern recognition problem. Simulation studies indicate energy savings by ∼100× in comparison to a corresponding digital/analog CMOS neuron implementation.",
author = "Abhronil Sengupta and Yong Shim and Kaushik Roy",
year = "2016",
month = "12",
doi = "10.1109/TBCAS.2016.2525823",
language = "English (US)",
volume = "10",
pages = "1152--1160",
journal = "IEEE Transactions on Biomedical Circuits and Systems",
issn = "1932-4545",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "6",

}

TY - JOUR

T1 - Proposal for an all-spin artificial neural network

T2 - Emulating neural and synaptic functionalities through domain wall motion in ferromagnets

AU - Sengupta, Abhronil

AU - Shim, Yong

AU - Roy, Kaushik

PY - 2016/12

Y1 - 2016/12

N2 - Non-Boolean computing based on emerging post-CMOS technologies can potentially pave the way for low-power neural computing platforms. However, existing work on such emerging neuromorphic architectures have either focused on solely mimicking the neuron, or the synapse functionality. While memristive devices have been proposed to emulate biological synapses, spintronic devices have proved to be efficient at performing the thresholding operation of the neuron at ultra-low currents. In this work, we propose an All-Spin Artificial Neural Network where a single spintronic device acts as the basic building block of the system. The device offers a direct mapping to synapse and neuron functionalities in the brain while inter-layer network communication is accomplished via CMOS transistors. To the best of our knowledge, this is the first demonstration of a neural architecture where a single nanoelectronic device is able to mimic both neurons and synapses. The ultra-low voltage operation of low resistance magneto-metallic neurons enables the low-voltage operation of the array of spintronic synapses, thereby leading to ultra-low power neural architectures. Device-level simulations, calibrated to experimental results, was used to drive the circuit and system level simulations of the neural network for a standard pattern recognition problem. Simulation studies indicate energy savings by ∼100× in comparison to a corresponding digital/analog CMOS neuron implementation.

AB - Non-Boolean computing based on emerging post-CMOS technologies can potentially pave the way for low-power neural computing platforms. However, existing work on such emerging neuromorphic architectures have either focused on solely mimicking the neuron, or the synapse functionality. While memristive devices have been proposed to emulate biological synapses, spintronic devices have proved to be efficient at performing the thresholding operation of the neuron at ultra-low currents. In this work, we propose an All-Spin Artificial Neural Network where a single spintronic device acts as the basic building block of the system. The device offers a direct mapping to synapse and neuron functionalities in the brain while inter-layer network communication is accomplished via CMOS transistors. To the best of our knowledge, this is the first demonstration of a neural architecture where a single nanoelectronic device is able to mimic both neurons and synapses. The ultra-low voltage operation of low resistance magneto-metallic neurons enables the low-voltage operation of the array of spintronic synapses, thereby leading to ultra-low power neural architectures. Device-level simulations, calibrated to experimental results, was used to drive the circuit and system level simulations of the neural network for a standard pattern recognition problem. Simulation studies indicate energy savings by ∼100× in comparison to a corresponding digital/analog CMOS neuron implementation.

UR - http://www.scopus.com/inward/record.url?scp=84969581374&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84969581374&partnerID=8YFLogxK

U2 - 10.1109/TBCAS.2016.2525823

DO - 10.1109/TBCAS.2016.2525823

M3 - Article

C2 - 27214912

AN - SCOPUS:84969581374

VL - 10

SP - 1152

EP - 1160

JO - IEEE Transactions on Biomedical Circuits and Systems

JF - IEEE Transactions on Biomedical Circuits and Systems

SN - 1932-4545

IS - 6

M1 - 7470633

ER -