MicroLEDs for optical neuromorphic computing: Application potential and present challenges

Publikation: Arbeitspapier/PreprintPreprint

Autoren

  • Robert Kraneis
  • Maximilian Müller
  • Steffen Higgins-Wood
  • Noah Kälin
  • Hergo Heinrich Wehmann
  • Norwin von Malm
  • Christian Werner
  • Andreas Waag

Externe Organisationen

  • Technische Universität Braunschweig
  • OST – Ostschweizer Fachhochschule
  • OSRAM Licht AG
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Seitenumfang13
PublikationsstatusElektronisch veröffentlicht (E-Pub) - 11 Juni 2024
Extern publiziertJa

Abstract

The slow non-radiative surface recombination velocity of gallium nitride (GaN) in combination with its highly efficient radiative recombination makes this material ideally suited for microLEDs with dimensions as small as 1 μm and even below, serving as the fundamental building block of micro-displays. However, due to their superior miniaturization potential and energy efficiency, GaN-based microLEDs have applications that extend well beyond display technology. Their capability to produce optical patterns with high resolution, which can be modulated at extremely high frequencies, makes them suitable for numerous other applications. We suggest exploiting these exciting properties for a new and potentially equally significant application: utilizing microLEDs in optical processing units for artificial intelligence workloads. In neuromorphic computing, relevant aspects of biological neural networks are emulated directly with either electronic circuits or photonic devices, avoiding the shortcomings of conventional digital computer technology for AI workloads, which generally require massively parallel information processing. GaN microLEDs are discussed here as a promising enabling technology for optical neuromorphic processing units. We will demonstrate their potential to substantially decrease power consumption through massively parallel in-memory processing combined with efficient photon production and detection. A theoretical analysis of scalability and energy efficiency is provided. A macroscopic bench-top optical microLED demonstrator is presented, which experimentally proves the feasibility of our approach. Future potential and challenges associated with miniaturizing and scaling microLED based optical processing units are discussed. Finally, we summarize the open research questions that require attention before fully functional and miniaturized optical neuromorphic processing units based on GaN microLEDs can be realized

Ziele für nachhaltige Entwicklung

Zitieren

MicroLEDs for optical neuromorphic computing: Application potential and present challenges. / Kraneis, Robert; Müller, Maximilian; Higgins-Wood, Steffen et al.
2024.

Publikation: Arbeitspapier/PreprintPreprint

Kraneis, R, Müller, M, Higgins-Wood, S, Kälin, N, Wehmann, HH, von Malm, N, Werner, C & Waag, A 2024 'MicroLEDs for optical neuromorphic computing: Application potential and present challenges'. https://doi.org/10.24355/dbbs.084-202406111128-0
Kraneis, R., Müller, M., Higgins-Wood, S., Kälin, N., Wehmann, H. H., von Malm, N., Werner, C., & Waag, A. (2024). MicroLEDs for optical neuromorphic computing: Application potential and present challenges. Vorabveröffentlichung online. https://doi.org/10.24355/dbbs.084-202406111128-0
Kraneis R, Müller M, Higgins-Wood S, Kälin N, Wehmann HH, von Malm N et al. MicroLEDs for optical neuromorphic computing: Application potential and present challenges. 2024 Jun 11. Epub 2024 Jun 11. doi: 10.24355/dbbs.084-202406111128-0
Kraneis, Robert ; Müller, Maximilian ; Higgins-Wood, Steffen et al. / MicroLEDs for optical neuromorphic computing : Application potential and present challenges. 2024.
Download
@techreport{7b41da95f3c14554bfa7849fcbd5f202,
title = "MicroLEDs for optical neuromorphic computing: Application potential and present challenges",
abstract = " The slow non-radiative surface recombination velocity of gallium nitride (GaN) in combination with its highly efficient radiative recombination makes this material ideally suited for microLEDs with dimensions as small as 1 μm and even below, serving as the fundamental building block of micro-displays. However, due to their superior miniaturization potential and energy efficiency, GaN-based microLEDs have applications that extend well beyond display technology. Their capability to produce optical patterns with high resolution, which can be modulated at extremely high frequencies, makes them suitable for numerous other applications. We suggest exploiting these exciting properties for a new and potentially equally significant application: utilizing microLEDs in optical processing units for artificial intelligence workloads. In neuromorphic computing, relevant aspects of biological neural networks are emulated directly with either electronic circuits or photonic devices, avoiding the shortcomings of conventional digital computer technology for AI workloads, which generally require massively parallel information processing. GaN microLEDs are discussed here as a promising enabling technology for optical neuromorphic processing units. We will demonstrate their potential to substantially decrease power consumption through massively parallel in-memory processing combined with efficient photon production and detection. A theoretical analysis of scalability and energy efficiency is provided. A macroscopic bench-top optical microLED demonstrator is presented, which experimentally proves the feasibility of our approach. Future potential and challenges associated with miniaturizing and scaling microLED based optical processing units are discussed. Finally, we summarize the open research questions that require attention before fully functional and miniaturized optical neuromorphic processing units based on GaN microLEDs can be realized",
author = "Robert Kraneis and Maximilian M{\"u}ller and Steffen Higgins-Wood and Noah K{\"a}lin and Wehmann, {Hergo Heinrich} and {von Malm}, Norwin and Christian Werner and Andreas Waag",
year = "2024",
month = jun,
day = "11",
doi = "10.24355/dbbs.084-202406111128-0",
language = "English",
type = "WorkingPaper",

}

Download

TY - UNPB

T1 - MicroLEDs for optical neuromorphic computing

T2 - Application potential and present challenges

AU - Kraneis, Robert

AU - Müller, Maximilian

AU - Higgins-Wood, Steffen

AU - Kälin, Noah

AU - Wehmann, Hergo Heinrich

AU - von Malm, Norwin

AU - Werner, Christian

AU - Waag, Andreas

PY - 2024/6/11

Y1 - 2024/6/11

N2 - The slow non-radiative surface recombination velocity of gallium nitride (GaN) in combination with its highly efficient radiative recombination makes this material ideally suited for microLEDs with dimensions as small as 1 μm and even below, serving as the fundamental building block of micro-displays. However, due to their superior miniaturization potential and energy efficiency, GaN-based microLEDs have applications that extend well beyond display technology. Their capability to produce optical patterns with high resolution, which can be modulated at extremely high frequencies, makes them suitable for numerous other applications. We suggest exploiting these exciting properties for a new and potentially equally significant application: utilizing microLEDs in optical processing units for artificial intelligence workloads. In neuromorphic computing, relevant aspects of biological neural networks are emulated directly with either electronic circuits or photonic devices, avoiding the shortcomings of conventional digital computer technology for AI workloads, which generally require massively parallel information processing. GaN microLEDs are discussed here as a promising enabling technology for optical neuromorphic processing units. We will demonstrate their potential to substantially decrease power consumption through massively parallel in-memory processing combined with efficient photon production and detection. A theoretical analysis of scalability and energy efficiency is provided. A macroscopic bench-top optical microLED demonstrator is presented, which experimentally proves the feasibility of our approach. Future potential and challenges associated with miniaturizing and scaling microLED based optical processing units are discussed. Finally, we summarize the open research questions that require attention before fully functional and miniaturized optical neuromorphic processing units based on GaN microLEDs can be realized

AB - The slow non-radiative surface recombination velocity of gallium nitride (GaN) in combination with its highly efficient radiative recombination makes this material ideally suited for microLEDs with dimensions as small as 1 μm and even below, serving as the fundamental building block of micro-displays. However, due to their superior miniaturization potential and energy efficiency, GaN-based microLEDs have applications that extend well beyond display technology. Their capability to produce optical patterns with high resolution, which can be modulated at extremely high frequencies, makes them suitable for numerous other applications. We suggest exploiting these exciting properties for a new and potentially equally significant application: utilizing microLEDs in optical processing units for artificial intelligence workloads. In neuromorphic computing, relevant aspects of biological neural networks are emulated directly with either electronic circuits or photonic devices, avoiding the shortcomings of conventional digital computer technology for AI workloads, which generally require massively parallel information processing. GaN microLEDs are discussed here as a promising enabling technology for optical neuromorphic processing units. We will demonstrate their potential to substantially decrease power consumption through massively parallel in-memory processing combined with efficient photon production and detection. A theoretical analysis of scalability and energy efficiency is provided. A macroscopic bench-top optical microLED demonstrator is presented, which experimentally proves the feasibility of our approach. Future potential and challenges associated with miniaturizing and scaling microLED based optical processing units are discussed. Finally, we summarize the open research questions that require attention before fully functional and miniaturized optical neuromorphic processing units based on GaN microLEDs can be realized

U2 - 10.24355/dbbs.084-202406111128-0

DO - 10.24355/dbbs.084-202406111128-0

M3 - Preprint

BT - MicroLEDs for optical neuromorphic computing

ER -