Towards a Robust Multiply-Accumulate Cell in Photonics using Phase-Change Materials

被引：1

作者：

Cardoso, Raphael ^{[1
]}

Zrounba, Clement ^{[1
]}

Abdalla, Mohab ^{[1
,2
]}

Jimenez, Paul ^{[1
]}

de Queiroz, Mauricio Gomes ^{[1
]}

Charbonnier, Benoit ^{[3
]}

Pavanello, Fabio ^{[1
]}

O'Connor, Ian ^{[1
]}

Le Beux, Sebastien ^{[4
]}

机构：

[1] Univ Lyon, CNRS, Ecole Centrale Lyon, INSA Lyon,CPE Lyon,INL,UMR5270, Ecully, France

[2] RMIT Univ, Sch Engn, Melbourne, Vic 3000, Australia

[3] CEA LETI Grenoble, Grenoble, France

[4] Concordia Univ, Dept Elect & Comp Engn, Montreal, PQ, Canada

来源：

2023 DESIGN, AUTOMATION & TEST IN EUROPE CONFERENCE & EXHIBITION, DATE | 2023年

关键词：

photonic computing; phase-change memories; stochastic computing; computation-in-memory;

D O I：

10.23919/DATE56975.2023.10137313

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

In this paper we propose a novel approach to multiply-accumulate (MAC) operations in photonics. This approach is based on stochastic computing and on the dynamic behavior of phase-change materials (PCMs), leading to the unique characteristic of automatically storing the result in nonvolatile memory. We demonstrate that, even with perfect look-up tables, the standard approach to PCM scalar multiplication is highly susceptible to perturbations as small as 0.1% of the input power, causing repetitive peaks of 600% relative error. In the same operating conditions, the proposed method achieves an average of 7x improvement in precision.

引用

页数：2

共 6 条

[1] Survey of Stochastic Computing [J].

Alaghi, Armin ;

Hayes, John P. .

ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS, 2013, 12

[2] Parallel convolutional processing using an integrated photonic tensor core [J].

Feldmann, J. ;

Youngblood, N. ;

Karpov, M. ;

Gehring, H. ;

Li, X. ;

Stappers, M. ;

Le Gallo, M. ;

Fu, X. ;

Lukashchuk, A. ;

Raja, A. S. ;

Liu, J. ;

Wright, C. D. ;

Sebastian, A. ;

Kippenberg, T. J. ;

Pernice, W. H. P. ;

Bhaskaran, H. .

NATURE, 2021, 589 (7840) :52-+

[3] Calculating with light using a chip-scale all-optical abacus [J].

Feldmann, J. ;

Stegmaier, M. ;

Gruhler, N. ;

Rios, C. ;

Bhaskaran, H. ;

Wright, C. D. ;

Pernice, W. H. P. .

NATURE COMMUNICATIONS, 2017, 8

[4]

Peserico N, 2022, Arxiv, DOI [arXiv:2211.01476, DOI 10.1109/JLT.2023.3269957]

[5] Programmable chalcogenide-based all-optical deep neural networks [J].

Teo, Ting Yu ;

Ma, Xiaoxuan ;

Pastor, Ernest ;

Wang, Hao ;

George, Jonathan K. ;

Yang, Joel K. W. ;

Wall, Simon ;

Miscuglio, Mario ;

Simpson, Robert E. ;

Sorger, Volker J. .

NANOPHOTONICS, 2022, 11 (17) :4073-4088

[6] Programmable phase-change metasurfaces on waveguides for multimode photonic convolutional neural network [J].

Wu, Changming ;

Yu, Heshan ;

Lee, Seokhyeong ;

Peng, Ruoming ;

Takeuchi, Ichiro ;

Li, Mo .

NATURE COMMUNICATIONS, 2021, 12 (01)

← 1 →