Comprehensive performance analysis of a VCSEL-based photonic reservoir computer

Julian Bueno; Joshua Robertson; Matěj Hejda; Antonio Hurtado

doi:10.1109/LPT.2021.3075095

Comprehensive performance analysis of a VCSEL-based photonic reservoir computer

Julian Bueno, Joshua Robertson, Matěj Hejda, Antonio Hurtado

Research output: Contribution to journal › Article › peer-review

20 Citations (Scopus)

72 Downloads (Pure)

Abstract

Optical neural networks offer radically new avenues for ultrafast, energy-efficient hardware for machine learning and artificial intelligence. Reservoir Computing (RC), given its high performance and cheap training has attracted considerable attention for photonic neural network implementations, principally based on semiconductor lasers (SLs). Among SLs, Vertical Cavity Surface Emitting Lasers (VCSELs) possess unique attributes, e.g. high speed, low power, rich dynamics, reduced cost, ease to integrate in array architectures, making them valuable candidates for future photonic neural networks. This work provides a comprehensive analysis of a telecom-wavelength GHz-rate VCSEL RC system, revealing the impact of key system parameters on its performance across different processing tasks.

Original language	English
Article number	9415868
Pages (from-to)	920-923
Number of pages	4
Journal	IEEE Photonics Technology Letters
Volume	33
Issue number	16
Early online date	26 Apr 2021
DOIs	https://doi.org/10.1109/LPT.2021.3075095
Publication status	Published - 15 Aug 2021

Keywords

optical computing
vertical cavity surface 15 emitting lasers
neural networks

Access to Document

10.1109/LPT.2021.3075095

Bueno-etal-ieeeptl2021- comprehensive-performance-analysis-of-a-vcsel-based-photonic-reservoir-computerAccepted author manuscript, 484 KBLicence: CC BY-NC 4.0

1 Active
3 Finished

Turing AI Fellowship: PHOTONics for ultrafast Artificial Intelligence
Hurtado, A.
EPSRC (Engineering and Physical Sciences Research Council)
1/01/21 → 31/12/25
Project: Research Fellowship
Energy-efficient and high-bandwidth neuromorphic nanophotonic chips for artificial intelligence systems (ChipAI) H2020-FETOPEN
Hurtado, A. & Strain, M.
European Commission - Horizon 2020
1/03/19 → 28/02/22
Project: Research
BRAIN-inspired networks of ultrafast LASER neurons
Hurtado, A.
Office of Naval Research ONR
10/11/17 → 9/11/21
Project: Research

Data for: "Comprehensive Performance Analysis of a VCSEL-based Photonic Reservoir Computer"
Bueno, J. (Creator), Robertson, J. (Contributor), Hejda, M. (Contributor) & Hurtado, A. (Supervisor), University of Strathclyde, 17 May 2021
DOI: 10.15129/8ef280c8-3e3c-4287-82f9-4aec45f79644
Dataset

Cite this

@article{d79bbfd7624c4583b5ac22f6a2f64734,

title = "Comprehensive performance analysis of a VCSEL-based photonic reservoir computer",

abstract = "Optical neural networks offer radically new avenues for ultrafast, energy-efficient hardware for machine learning and artificial intelligence. Reservoir Computing (RC), given its high performance and cheap training has attracted considerable attention for photonic neural network implementations, principally based on semiconductor lasers (SLs). Among SLs, Vertical Cavity Surface Emitting Lasers (VCSELs) possess unique attributes, e.g. high speed, low power, rich dynamics, reduced cost, ease to integrate in array architectures, making them valuable candidates for future photonic neural networks. This work provides a comprehensive analysis of a telecom-wavelength GHz-rate VCSEL RC system, revealing the impact of key system parameters on its performance across different processing tasks. ",

keywords = "optical computing, vertical cavity surface 15 emitting lasers, neural networks",

author = "Julian Bueno and Joshua Robertson and Mat{\v e}j Hejda and Antonio Hurtado",

year = "2021",

month = aug,

day = "15",

doi = "10.1109/LPT.2021.3075095",

language = "English",

volume = "33",

pages = "920--923",

journal = "IEEE Photonics Technology Letters",

issn = "1041-1135",

number = "16",

}

TY - JOUR

T1 - Comprehensive performance analysis of a VCSEL-based photonic reservoir computer

AU - Bueno, Julian

AU - Robertson, Joshua

AU - Hejda, Matěj

AU - Hurtado, Antonio

PY - 2021/8/15

Y1 - 2021/8/15

N2 - Optical neural networks offer radically new avenues for ultrafast, energy-efficient hardware for machine learning and artificial intelligence. Reservoir Computing (RC), given its high performance and cheap training has attracted considerable attention for photonic neural network implementations, principally based on semiconductor lasers (SLs). Among SLs, Vertical Cavity Surface Emitting Lasers (VCSELs) possess unique attributes, e.g. high speed, low power, rich dynamics, reduced cost, ease to integrate in array architectures, making them valuable candidates for future photonic neural networks. This work provides a comprehensive analysis of a telecom-wavelength GHz-rate VCSEL RC system, revealing the impact of key system parameters on its performance across different processing tasks.

AB - Optical neural networks offer radically new avenues for ultrafast, energy-efficient hardware for machine learning and artificial intelligence. Reservoir Computing (RC), given its high performance and cheap training has attracted considerable attention for photonic neural network implementations, principally based on semiconductor lasers (SLs). Among SLs, Vertical Cavity Surface Emitting Lasers (VCSELs) possess unique attributes, e.g. high speed, low power, rich dynamics, reduced cost, ease to integrate in array architectures, making them valuable candidates for future photonic neural networks. This work provides a comprehensive analysis of a telecom-wavelength GHz-rate VCSEL RC system, revealing the impact of key system parameters on its performance across different processing tasks.

KW - optical computing

KW - vertical cavity surface 15 emitting lasers

KW - neural networks

U2 - 10.1109/LPT.2021.3075095

DO - 10.1109/LPT.2021.3075095

M3 - Article

SN - 1041-1135

VL - 33

SP - 920

EP - 923

JO - IEEE Photonics Technology Letters

JF - IEEE Photonics Technology Letters

IS - 16

M1 - 9415868

ER -

Comprehensive performance analysis of a VCSEL-based photonic reservoir computer

Abstract

Keywords

Access to Document

Fingerprint

Projects

Turing AI Fellowship: PHOTONics for ultrafast Artificial Intelligence

Energy-efficient and high-bandwidth neuromorphic nanophotonic chips for artificial intelligence systems (ChipAI) H2020-FETOPEN

BRAIN-inspired networks of ultrafast LASER neurons

Datasets

Data for: "Comprehensive Performance Analysis of a VCSEL-based Photonic Reservoir Computer"

Cite this