Prospects and applications of photonic neural networks

被引:127
作者
Huang, Chaoran [1 ,2 ]
Sorger, Volker J. [3 ]
Miscuglio, Mario [3 ]
Al-Qadasi, Mohammed
Mukherjee, Avilash [4 ]
Lampe, Lutz [4 ]
Nichols, Mitchell [4 ]
Tait, Alexander N. [5 ]
Ferreira de Lima, Thomas [1 ]
Marquez, Bicky A. [6 ]
Wang, Jiahui [7 ]
Chrostowski, Lukas [4 ]
Fok, Mable P. [8 ]
Brunner, Daniel [9 ]
Fan, Shanhui [7 ]
Shekhar, Sudip [4 ]
Prucnal, Paul R. [1 ]
Shastri, Bhavin J. [6 ,10 ]
机构
[1] Princeton Univ, Dept Elect & Comp Engn, Princeton, NJ 08544 USA
[2] Chinese Univ Hong Kong, Dept Elect Engn, Hong Kong, Peoples R China
[3] George Washington Univ, Dept Elect & Comp Engn, Washington, DC USA
[4] Univ British Columbia, Dept Elect & Comp Engn, Vancouver, BC, Canada
[5] NIST, Boulder, CO USA
[6] Queens Univ, Dept Elect Engn, Kingston, ON, Canada
[7] Stanford Univ, Coll Engn, Stanford, CA 94305 USA
[8] Univ Georgia, Athens, GA USA
[9] Univ Bourgogne Franche Comte, Inst FEMTO ST, Dept Phys Engn Phys & Astron, CNRS,UMR, Besancon, France
[10] Vector Inst, Toronto, ON, Canada
基金
加拿大自然科学与工程研究理事会;
关键词
Photonic neural networks; neuromorphic photonics; silicon photonics; neuromorphic computing; machine learning; artificial intelligence; MACH-ZEHNDER MODULATOR; OPTICAL FLIP-FLOP; SAR ADC; ARTIFICIAL-INTELLIGENCE; ULTRA-COMPACT; DESIGN; ENERGY; ANALOG; BACKPROPAGATION; IMPLEMENTATION;
D O I
10.1080/23746149.2021.1981155
中图分类号
O4 [物理学];
学科分类号
0702 ;
摘要
Neural networks have enabled applications in artificial intelligence through machine learning, and neuromorphic computing. Software implementations of neural networks on conventional computers that have separate memory and processor (and that operate sequentially) are limited in speed and energy efficiency. Neuromorphic engineering aims to build processors in which hardware mimics neurons and synapses in the brain for distributed and parallel processing. Neuromorphic engineering enabled by photonics (optical physics) can offer sub-nanosecond latencies and high bandwidth with low energies to extend the domain of artificial intelligence and neuromorphic computing applications to machine learning acceleration, nonlinear programming, intelligent signal processing, etc. Photonic neural networks have been demonstrated on integrated platforms and free-space optics depending on the class of applications being targeted. Here, we discuss the prospects and demonstrated applications of these photonic neural networks.
引用
收藏
页数:63
相关论文
共 203 条
[1]   Designing and Modeling Analog Neural Network Training Accelerators [J].
Agarwal, Sapan ;
Jacobs-Gedrim, Robin B. ;
Bennett, Christopher ;
Hsia, Alex ;
Van Heukelom, Michael S. ;
Hughart, David ;
Fuller, Elliot ;
Li, Yiyang ;
Talin, A. Alec ;
Marinella, Matthew J. .
2019 INTERNATIONAL SYMPOSIUM ON VLSI TECHNOLOGY, SYSTEMS AND APPLICATION (VLSI-TSA), 2019,
[2]   Roadmap of optical communications [J].
Agrell, Erik ;
Karlsson, Magnus ;
Chraplyvy, A. R. ;
Richardson, David J. ;
Krummrich, Peter M. ;
Winzer, Peter ;
Roberts, Kim ;
Fischer, Johannes Karl ;
Savory, Seb J. ;
Eggleton, Benjamin J. ;
Secondini, Marco ;
Kschischang, Frank R. ;
Lord, Andrew ;
Prat, Josep ;
Tomkos, Ioannis ;
Bowers, John E. ;
Srinivasan, Sudha ;
Brandt-Pearce, Maite ;
Gisin, Nicolas .
JOURNAL OF OPTICS, 2016, 18 (06)
[3]   A Dual-Polarization Silicon-Photonic Coherent Transmitter Supporting 552 Gb/s/wavelength [J].
Ahmed, Abdelrahman H. ;
El Moznine, Abdellatif ;
Lim, Daihyun ;
Ma, Yangjin ;
Rylyakov, Alexander ;
Shekhar, Sudip .
IEEE JOURNAL OF SOLID-STATE CIRCUITS, 2020, 55 (09) :2597-2608
[4]   Silicon-Photonics Microring Links for Datacenters-Challenges and Opportunities [J].
Ahmed, Abdelrahman H. ;
Sharkia, Ahmad ;
Casper, Bryan ;
Mirabbasi, Shahriar ;
Shekhar, Sudip .
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, 2016, 22 (06) :194-203
[5]  
Akrout M., 2019, Deep Learning without Weight Transport
[6]   Optical RAM and integrated optical memories: a survey [J].
Alexoudi, Theoni ;
Kanellos, George Theodore ;
Pleros, Nikos .
LIGHT-SCIENCE & APPLICATIONS, 2020, 9 (01)
[7]  
Ambs Pierre, 2010, Advances in Optical Technologies, DOI 10.1155/2010/372652
[8]   ITO-based electro-absorption modulator for photonic neural activation function [J].
Amin, R. ;
George, J. K. ;
Sun, S. ;
de Lima, T. Ferreira ;
Tait, A. N. ;
Khurgin, J. B. ;
Miscuglio, M. ;
Shastri, B. J. ;
Prucnal, P. R. ;
El-Ghazawi, T. ;
Sorger, V. J. .
APL MATERIALS, 2019, 7 (08)
[9]   Sub-wavelength GHz-fast broadband ITO Mach-Zehnder modulator on silicon photonics [J].
Amin, Rubab ;
Maiti, Rishi ;
Gui, Yaliang ;
Suer, Can ;
Miscuglio, Mario ;
Heidari, Elham ;
Chen, Ray T. ;
Dalir, Hamed ;
Sorger, Volker J. .
OPTICA, 2020, 7 (04) :333-335
[10]   A Lateral MOS-Capacitor-Enabled ITO Mach-Zehnder Modulator for Beam Steering [J].
Amin, Rubab ;
Maiti, Rishi ;
George, Jonathan K. ;
Ma, Xiaoxuan ;
Ma, Zhizhen ;
Dalir, Hamed ;
Miscuglio, Mario ;
Sorger, Volker J. .
JOURNAL OF LIGHTWAVE TECHNOLOGY, 2020, 38 (02) :282-290