PIXEL: Photonic neural network accelerator

Kyle Shiflett, Dylan Wright, Avinash Karanth, Ahmed Louri

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

15 Scopus citations

Abstract

Machine learning (ML) architectures such as Deep Neural Networks (DNNs) have achieved unprecedented accuracy on modern applications such as image classification and speech recognition. With power dissipation becoming a major concern in ML architectures, computer architects have focused on designing both energy-efficient hardware platforms as well as optimizing ML algorithms. To dramatically reduce power consumption and increase parallelism in neural network accelerators, disruptive technology such as silicon photonics has been proposed which can improve the performance-per-Watt when compared to electrical implementation. In this paper, we propose PIXEL - Photonic Neural Network Accelerator that efficiently implements the fundamental operation in neural computation, namely the multiply and accumulate (MAC) functionality using photonic components such as microring resonators (MRRs) and Mach-Zehnder interferometer (MZI). We design two versions of PIXEL - a hybrid version that multiplies optically and accumulates electrically and a fully optical version that multiplies and accumulates optically. We perform a detailed power, area and timing analysis of the different versions of photonic and electronic accelerators for different convolution neural networks (AlexNet, VGG16, and others). Our results indicate a significant improvement in the energy-delay product for both PIXEL designs over traditional electrical designs (48.4% for OE and 73.9% for OO) while minimizing latency, at the cost of increased area over electrical designs.

Original languageEnglish (US)
Title of host publicationProceedings - 2020 IEEE International Symposium on High Performance Computer Architecture, HPCA 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages474-487
Number of pages14
ISBN (Electronic)9781728161495
DOIs
StatePublished - Feb 2020
Event26th IEEE International Symposium on High Performance Computer Architecture, HPCA 2020 - San Diego, United States
Duration: Feb 22 2020Feb 26 2020

Publication series

NameProceedings - 2020 IEEE International Symposium on High Performance Computer Architecture, HPCA 2020

Conference

Conference26th IEEE International Symposium on High Performance Computer Architecture, HPCA 2020
Country/TerritoryUnited States
CitySan Diego
Period2/22/202/26/20

Keywords

  • Accelerator
  • Deep neural network
  • Mach-Zehnder interferometer
  • Machine learning
  • Microring resonator
  • Silicon photonics

ASJC Scopus subject areas

  • Artificial Intelligence
  • Hardware and Architecture
  • Safety, Risk, Reliability and Quality

Fingerprint

Dive into the research topics of 'PIXEL: Photonic neural network accelerator'. Together they form a unique fingerprint.

Cite this