Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons

Indranil Chakraborty, Gobinda Saha, Abhronil Sengupta, Kaushik Roy

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons and synapses. However, efforts have largely been invested in implementations in the electrical domain with potential limitations of switching speed, packing density of large integrated systems and interconnect losses. As an alternative, neuromorphic engineering in the photonic domain has recently gained attention. In this work, we propose a purely photonic operation of an Integrate-and-Fire Spiking neuron, based on the phase change dynamics of Ge2Sb2Te5 (GST) embedded on top of a microring resonator, which alleviates the energy constraints of PCMs in electrical domain. We also show that such a neuron can be potentially integrated with on-chip synapses into an all-Photonic Spiking Neural network inferencing framework which promises to be ultrafast and can potentially offer a large operating bandwidth.

Original languageEnglish (US)
Article number12980
JournalScientific reports
Volume8
Issue number1
DOIs
StatePublished - Dec 1 2018

Fingerprint

Optics and Photonics
Neurons
Synapses
Brain
Growth

All Science Journal Classification (ASJC) codes

  • General

Cite this

Chakraborty, Indranil ; Saha, Gobinda ; Sengupta, Abhronil ; Roy, Kaushik. / Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons. In: Scientific reports. 2018 ; Vol. 8, No. 1.
@article{9e2326b8384f4b06a60d2d2263238129,
title = "Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons",
abstract = "The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons and synapses. However, efforts have largely been invested in implementations in the electrical domain with potential limitations of switching speed, packing density of large integrated systems and interconnect losses. As an alternative, neuromorphic engineering in the photonic domain has recently gained attention. In this work, we propose a purely photonic operation of an Integrate-and-Fire Spiking neuron, based on the phase change dynamics of Ge2Sb2Te5 (GST) embedded on top of a microring resonator, which alleviates the energy constraints of PCMs in electrical domain. We also show that such a neuron can be potentially integrated with on-chip synapses into an all-Photonic Spiking Neural network inferencing framework which promises to be ultrafast and can potentially offer a large operating bandwidth.",
author = "Indranil Chakraborty and Gobinda Saha and Abhronil Sengupta and Kaushik Roy",
year = "2018",
month = "12",
day = "1",
doi = "10.1038/s41598-018-31365-x",
language = "English (US)",
volume = "8",
journal = "Scientific Reports",
issn = "2045-2322",
publisher = "Nature Publishing Group",
number = "1",

}

Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons. / Chakraborty, Indranil; Saha, Gobinda; Sengupta, Abhronil; Roy, Kaushik.

In: Scientific reports, Vol. 8, No. 1, 12980, 01.12.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons

AU - Chakraborty, Indranil

AU - Saha, Gobinda

AU - Sengupta, Abhronil

AU - Roy, Kaushik

PY - 2018/12/1

Y1 - 2018/12/1

N2 - The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons and synapses. However, efforts have largely been invested in implementations in the electrical domain with potential limitations of switching speed, packing density of large integrated systems and interconnect losses. As an alternative, neuromorphic engineering in the photonic domain has recently gained attention. In this work, we propose a purely photonic operation of an Integrate-and-Fire Spiking neuron, based on the phase change dynamics of Ge2Sb2Te5 (GST) embedded on top of a microring resonator, which alleviates the energy constraints of PCMs in electrical domain. We also show that such a neuron can be potentially integrated with on-chip synapses into an all-Photonic Spiking Neural network inferencing framework which promises to be ultrafast and can potentially offer a large operating bandwidth.

AB - The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons and synapses. However, efforts have largely been invested in implementations in the electrical domain with potential limitations of switching speed, packing density of large integrated systems and interconnect losses. As an alternative, neuromorphic engineering in the photonic domain has recently gained attention. In this work, we propose a purely photonic operation of an Integrate-and-Fire Spiking neuron, based on the phase change dynamics of Ge2Sb2Te5 (GST) embedded on top of a microring resonator, which alleviates the energy constraints of PCMs in electrical domain. We also show that such a neuron can be potentially integrated with on-chip synapses into an all-Photonic Spiking Neural network inferencing framework which promises to be ultrafast and can potentially offer a large operating bandwidth.

UR - http://www.scopus.com/inward/record.url?scp=85052391736&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052391736&partnerID=8YFLogxK

U2 - 10.1038/s41598-018-31365-x

DO - 10.1038/s41598-018-31365-x

M3 - Article

C2 - 30154507

AN - SCOPUS:85052391736

VL - 8

JO - Scientific Reports

JF - Scientific Reports

SN - 2045-2322

IS - 1

M1 - 12980

ER -