Lightspeed Binary Neural Networks using Optical Phase-Change Materials

Conference paper (2023)

Authors

Taha Shahroodi Computer Engineering - EEMCS
Rafaela Cardoso École Centrale de Lyon
M.Z. Zahedi Computer Engineering - EEMCS
J.S.S.M. Wong Computer Engineering - EEMCS
Alberto Bosio École Centrale de Lyon
Ian O'Connor École Centrale de Lyon
Said Hamdioui Quantum & Computer Engineering
S. Hamdioui Quantum & Computer Engineering
Research Group
Computer Engineering (EEMCS) (TU Delft)
Copyright: © 2023 T. Shahroodi, Rafaela Cardoso, M.Z. Zahedi, J.S.S.M. Wong, Alberto Bosio, Ian O'Connor, S. Hamdioui
DOI: https://doi.org/10.23919/DATE56975.2023.10137229
More Info
expand_more

Published Date

2023

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Department

Quantum & Computer Engineering

Research Group

Computer Engineering

Abstract

This paper investigates the potential of a compute-in-memory core based on optical Phase Change Materials (oPCMs) to speed up and reduce the energy consumption of the Matrix-Matrix-Multiplication operation. The paper also proposes a new data mapping for Binary Neural Networks (BNNs) tailored for our oPCM core. The preliminary results show a significant latency improvement irrespective of the evaluated network structure and size. The improvement varies from network to network and goes up to ~1053x.

Files

Lightspeed_Binary_Neural_Netwo... (pdf)
(pdf | 0.407 Mb)
- Embargo expired in 02-12-2023
Unknown license