Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse

Beomsang Yoo, Kiryong Kim, Seongook Jung

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Emerging nonvolatile memory technologies have been much studied to realize massively interconnected Spiking Neural Networks (SNNs) due to its high density, and energy efficiency. Nevertheless, most of the previous studies on utilizing such technologies have focused on implementing a traditional pair-based Spike Timing Dependent Plasticity (STDP) rule, which turned out to fail to reproduce multiple physiological experimental results. In this paper, we present a neuromorphic circuit for a three-terminal spintronic synapse with a triplet-based STDP rule, which is a higher order of learning mechanism. Simulation results indicate that the proposed learning circuit for the triplet-based STDP rule significantly improves learning capability in mimicking the various biological experimental data. We introduce a normalized mean-square error E value to evaluate the performance of each of the learning circuits quantitatively. The proposed learning circuit achieves E value of 1.77, which is far better than the conventional one that achieves E value of 12.2.

Original languageEnglish
Title of host publication2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages689-692
Number of pages4
ISBN (Electronic)9781538695623
DOIs
Publication statusPublished - 2019 Jan 17
Event25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018 - Bordeaux, France
Duration: 2018 Dec 92018 Dec 12

Publication series

Name2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018

Conference

Conference25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018
CountryFrance
CityBordeaux
Period18/12/918/12/12

Fingerprint

synapses
Magnetoelectronics
spikes
plastic properties
learning
Plasticity
time measurement
Networks (circuits)
spiking
Mean square error
Energy efficiency
emerging
Neural networks
Data storage equipment
simulation

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Instrumentation

Cite this

Yoo, B., Kim, K., & Jung, S. (2019). Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse. In 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018 (pp. 689-692). [8617974] (2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICECS.2018.8617974
Yoo, Beomsang ; Kim, Kiryong ; Jung, Seongook. / Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse. 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 689-692 (2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018).
@inproceedings{96fe876ed0644922915decf65315caf0,
title = "Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse",
abstract = "Emerging nonvolatile memory technologies have been much studied to realize massively interconnected Spiking Neural Networks (SNNs) due to its high density, and energy efficiency. Nevertheless, most of the previous studies on utilizing such technologies have focused on implementing a traditional pair-based Spike Timing Dependent Plasticity (STDP) rule, which turned out to fail to reproduce multiple physiological experimental results. In this paper, we present a neuromorphic circuit for a three-terminal spintronic synapse with a triplet-based STDP rule, which is a higher order of learning mechanism. Simulation results indicate that the proposed learning circuit for the triplet-based STDP rule significantly improves learning capability in mimicking the various biological experimental data. We introduce a normalized mean-square error E value to evaluate the performance of each of the learning circuits quantitatively. The proposed learning circuit achieves E value of 1.77, which is far better than the conventional one that achieves E value of 12.2.",
author = "Beomsang Yoo and Kiryong Kim and Seongook Jung",
year = "2019",
month = "1",
day = "17",
doi = "10.1109/ICECS.2018.8617974",
language = "English",
series = "2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "689--692",
booktitle = "2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018",
address = "United States",

}

Yoo, B, Kim, K & Jung, S 2019, Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse. in 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018., 8617974, 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018, Institute of Electrical and Electronics Engineers Inc., pp. 689-692, 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018, Bordeaux, France, 18/12/9. https://doi.org/10.1109/ICECS.2018.8617974

Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse. / Yoo, Beomsang; Kim, Kiryong; Jung, Seongook.

2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018. Institute of Electrical and Electronics Engineers Inc., 2019. p. 689-692 8617974 (2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse

AU - Yoo, Beomsang

AU - Kim, Kiryong

AU - Jung, Seongook

PY - 2019/1/17

Y1 - 2019/1/17

N2 - Emerging nonvolatile memory technologies have been much studied to realize massively interconnected Spiking Neural Networks (SNNs) due to its high density, and energy efficiency. Nevertheless, most of the previous studies on utilizing such technologies have focused on implementing a traditional pair-based Spike Timing Dependent Plasticity (STDP) rule, which turned out to fail to reproduce multiple physiological experimental results. In this paper, we present a neuromorphic circuit for a three-terminal spintronic synapse with a triplet-based STDP rule, which is a higher order of learning mechanism. Simulation results indicate that the proposed learning circuit for the triplet-based STDP rule significantly improves learning capability in mimicking the various biological experimental data. We introduce a normalized mean-square error E value to evaluate the performance of each of the learning circuits quantitatively. The proposed learning circuit achieves E value of 1.77, which is far better than the conventional one that achieves E value of 12.2.

AB - Emerging nonvolatile memory technologies have been much studied to realize massively interconnected Spiking Neural Networks (SNNs) due to its high density, and energy efficiency. Nevertheless, most of the previous studies on utilizing such technologies have focused on implementing a traditional pair-based Spike Timing Dependent Plasticity (STDP) rule, which turned out to fail to reproduce multiple physiological experimental results. In this paper, we present a neuromorphic circuit for a three-terminal spintronic synapse with a triplet-based STDP rule, which is a higher order of learning mechanism. Simulation results indicate that the proposed learning circuit for the triplet-based STDP rule significantly improves learning capability in mimicking the various biological experimental data. We introduce a normalized mean-square error E value to evaluate the performance of each of the learning circuits quantitatively. The proposed learning circuit achieves E value of 1.77, which is far better than the conventional one that achieves E value of 12.2.

UR - http://www.scopus.com/inward/record.url?scp=85062291596&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062291596&partnerID=8YFLogxK

U2 - 10.1109/ICECS.2018.8617974

DO - 10.1109/ICECS.2018.8617974

M3 - Conference contribution

T3 - 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018

SP - 689

EP - 692

BT - 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Yoo B, Kim K, Jung S. Triplet-based Spike Timing Dependent Plasticity Circuit Design for three-terminal Spintronic Synapse. In 2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018. Institute of Electrical and Electronics Engineers Inc. 2019. p. 689-692. 8617974. (2018 25th IEEE International Conference on Electronics Circuits and Systems, ICECS 2018). https://doi.org/10.1109/ICECS.2018.8617974