DOCSLIB.ORG

Artificial Intelligence: from Electronics to Optics

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Artificial Intelligence: from Electronics to Optics BACK TO BASICS ARTIFICIAL INTELLIGENCE: FROM ELECTRONICS TO OPTICS /////////////////////////////////////////////////////////////////////////////////////////////////// Sylvain GIGAN1,3,*, Florent KRZAKALA2,3, Laurent DAUDET3, Igor CARRON3 1 Laboratoire Kastler Brossel, ENS-Université PSL, CNRS, Sorbonne Université, Collège de France, Paris, France 2 Laboratoire de Physique de l’École Normale Supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université de Paris, Paris, France 3 LightOn, 2 rue de la Bourse, Paris, France *[email protected] Machine Learning and big data are currently revolutionizing our way of life, in particular with the recent emergence of deep learning. Powered by CPU and GPU, they are currently hardware limited and extremely energy intensive. Photonics, either integrated or in free space, offers a very promising alternative for realizing optically machine learning tasks at high speed and low consumption. We here review the history and current state of the art of optical computing and optical machine learning. https://doi.org/10.1051/photon/202010449 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ince the dawn of voice recognition to translation, the memory through a communi- micro-electronics image analysis to future self-driving cation bus. This architecture has and the emergence cars. However, machine learning’s been basically unchanged since its of lasers, both op- progress requires exponentially in- inception, and only improved thanks tics and electronics creasing resources, be it in memo- to the progress of microelectronics platforms have been ry, raw computing power, or energy and nanolithography, allowing the competing for infor- consumption. We introduce in this feature sizes of components to shrink mation processing and transmission. article the basics of neural networks, to 7 or less nanometers nowadays. While electronics has been overwhel- and see how this new architecture This has consequently diminished Smingly dominating computing for the shatters the status quo and provides tremendously the Ohmic losses and last 50 years thanks to Moore’s law, op- optics a new opportunity to shine in the energy consumption to a few pJ/ tics and photonics have been increa- computing, whether in free space or operation, and allowed the increase singly dominant for communications, in integrated photonics circuits. of the operating clock frequencies from long distance communications of the components to reach several with optical fibers to optical inter- EXPECTED CONTENTS GHz. Thus, component density has connects in data centers. Machine Optical computing. Classical com- driven the number of transistors on Learning, that also originated in the puting, such as the one running on a processor to several tens of billions, 1950s, has seen tremendous develop- our PC, is based on the so-called Von- while driving its cost down. This is ments in the last decade. The emer- Neuman architecture laid out in the the well-known Moore’s law, lea- gence of deep neural networks has 1940s, where a program is stored in ding to the observation that a good become the de facto standard for big a memory, and instructions are read desktop PC nowadays has a proces- data analysis and many of the tasks and executed on a processor, while sing power of several TeraFlops (1012 that we today consider normal: from input and output are exchanged in floating point operations per second). Photoniques 104 I www.photoniques.com 49 BACK TO BASICS Optics has several advantages Figure 1. The network also includes input and compared to electronics: its intrin- Some historical examples of optical computing. output neurons, that either receive or sic parallelism, its almost unlimited Left: the 1972 Tilted Plane Optical processor send information to and from the out- bandwidth, the ability of simple used for synthetic Aperture Radar all optical side world. Just like the brain, a neu- transformation by simple propaga- image processing (from Kosma et al. Applied ral network can be made to “learn”, tion (such as a Fourier Transform) Optics 11.8 (1972): 1766-1777), right, a vector- i.e. be optimized for a given task, by compared to electrons [1]. Thus, matrix-multiplier with optical feedback adjusting its weights, for instance optics has from the very start been (from Psaltis et al. Optics Letters 10.2 (1985): being fed at the input with images, considered as a viable alternative for 98-100) Reprinted with permission from being able to classify them into cate- analog computing. In the eighties, the © The Optical Society. gories. The analogy with the brain emergence of optical non-linearities, stops there: while the brain counts semi-conductor lasers and optical approximately 80 billions neurons memories has given hope that optics are connected to each other in very and 100 trillions connections, ANNs may be used to build an all-purpose complex networks, and where the have to be limited to much less neu- computing platform. Alas, the pro- response of a neuron can be trig- rons and weights, and to much sim- gress in optics failed to match Moore’s gered in a complex and non-linear pler architectures, in order to make law exponential pace, and the hope way by the electric influx it receives the training of the network possible. of building such an all-purpose opti- from many other neurons. Artificial Several typical architectures have cal computer was abandoned in the neural networks are similarly made been developed over the last decades, nineties [2]. Still, optics found nume- of “neurons” or nodes, that integrate to maximize efficiency on a given rous applications in storage, and of signals from other neurons, with va- task, while keeping the training of course in telecommunications, both rious weights, and emit a resulting the neural network computationally in long-distance with optical fibers, signal based on a non-linear activa- tractable. Most of the time, neurons and more recently in interconnects. tion function. This signal is, in turns, are organized in layers of various sizes Neural networks. In parallel, a fed to a number of other neurons. (number of neurons) and connecti- computing paradigm, resolutely vity. It ranges from the simplest different from conventional pro- networks, such as the perceptron gramming, emerged also in the Figure 2. (a single layer linking N inputs to a 1950’s: Artificial neural networks or Structure of an artificial neural network. Left: single output) which was one of the ANN, on which all modern artificial an artificial “neuron” comprising several inputs earliest ANN, to multi-layered fee- intelligence is based. It is (loosely) value, and one or many outputs, result of the dforward neural networks (where inspired from the structure and be- non-linear combinations of the inputs. Center neurons are organized in successive haviour of the brain, where neurons and left: two popular ANN architectures. layers and information is passed from 50 www.photoniques.com I Photoniques 104 BACK TO BASICS layer to layer) to recurrent networks GPU and CPUs. The rise of deep (where information can flow backwards learning and big data has been mostly and be fed back to previous layers). The powered by Moore’s law, allowing trai- connections from a layer to the next can ning and inference of very large neural be very sparse, in particular convolutio- networks. An important factor driving nal layers, or dense (all-to-all connected). deep learning is the transition to Graphic The performance of ANNs depends on Processing Units (GPU). Initially de- its structure, for instance a perceptron signed for computer graphics, these is good for simple linear classification, specialized processors were optimized but more complex tasks require more for parallel processing of large vectors complex network structures. and matrices. For neural networks, Deep learning. While artificial intel- where training and inference require ligence saw good progresses, until the a vast number of such multiplications, early 2000s, its overall performance for GPUs turned out to be much more power- day-to-day tasks remained modest and ful than CPUs (Central Processor Unit) did not find any clear real-life applica- and are now ubiquitous - incidentally, tions. This changed tremendously in the NVIDIA, the leader in GPU for deep lear- last two decades, thanks to the emer- ning, has now a capitalization that is on gence of a powerful architecture: Deep par with Intel. However, GPU and CPU learning, and its corollary networks, are still enormously power-hungry: it has known as Deep Neural networks. Deep been shown that training a single neu- Neural networks are layered networks ral network can use as much energy as with a large number of “hidden” layers. 5 cars over their lifetime, and more glo- Pioneers such as Yann Lecun, Yoshua bally, big data and data centers already Bengio and Geoffrey Hinton, have shown account for an estimated 4% of our en- that deep neural networks have an abi- ergy, and it may grow to over half of our lity to solve highly complex problems energy consumption in the next decade, [3]: in essence, while the first layers can if nothing changes. Meanwhile, Moore’s pre-process the input information (for law is officially stalling: nanolithography instance contours in images, or words and transistors are reaching their phy- in text), deeper layers can gradually sical limits, progresses in consumption distil more abstract concepts, such as and speed are getting much slower [4]. identifying an object, or extracting the Worse, the implementation of neural sense of a text. Nowadays, deep learning networks on both CPU and GPU suffers has demonstrated unprecedented per- from the so-called “Von Neumann bottle- formance at tasks that we only recently neck”: the bus transferring data between believed would be forever out of reach of memory and computing units ultimately machines, from beating the best player limits performance.
Load more

Recommended publications
  • Denys I. Bondar
    Updated on July 15, 2020 Denys I. Bondar Assistant Professor, Department of Physics and Engineering Physics, Tulane University, 2001 Percival Stern Hall, New Orleans, Louisiana, USA 70118 office: 4031 Percival Stern Hall e-mail: [email protected] phone: +1 (504) 862 8701 web: Google Scholar, Research Gate, ORCiD Education Ph. D. in Physics 01/2007{12/2010 Department of Physics and Astronomy, • University of Waterloo, Waterloo, Ontario, Canada Ph. D. thesis [arXiv:1012.5334]: supervisor: Misha Yu. Ivanov; co-supervisor: Wing-Ki Liu M. Sc. and B. Sc. in Physics with Honors 09/2001{06/2006 • Uzhgorod National University, Uzhgorod, Ukraine B. Sc. in Computer Science 09/2001{06/2006 • Transcarpathian State University, Uzhgorod, Ukraine Awards, Grants, and Scholarships Young Faculty Award DARPA 2019{2021 Army Research Office (ARO) grant W911NF-19-1-0377 2019{2021 Defense University Research Instrumentation Program (DURIP) 2018 Humboldt Research Fellowship for Experienced Researchers 2017-2020 Air Force Young Investigator Research Program 2016-2019 Los Alamos Director's Fellowship (declined) 2013 President's Graduate Scholarship (University of Waterloo) 2010 Ontario Graduate Scholarship 2010 International Doctoral Student Awards (University of Waterloo) 2007-2010 Award of Recognition at the All-Ukrainian Contest of Students' Scientific Works 2006 Current Research Interests • Quantum technology • Optics including quantum, ultrafast, nonlinear, and incoherent • Optical communication and sensing • Nonequilibrium quantum statistical mechanics 1 • Many-body
    [Show full text]
  • Nanosystems, Edge Computing, and the Next Generation Computing Systems
    sensors Review Nanosystems, Edge Computing, and the Next Generation Computing Systems Ali Passian * and Neena Imam Computing & Computational Sciences Directorate, Oak Ridge National Laboratory, Oak Ridge, TN 37830, USA; [email protected] * Correspondence: [email protected] Received: 30 July 2019; Accepted: 16 September 2019; Published: 19 September 2019 Abstract: It is widely recognized that nanoscience and nanotechnology and their subfields, such as nanophotonics, nanoelectronics, and nanomechanics, have had a tremendous impact on recent advances in sensing, imaging, and communication, with notable developments, including novel transistors and processor architectures. For example, in addition to being supremely fast, optical and photonic components and devices are capable of operating across multiple orders of magnitude length, power, and spectral scales, encompassing the range from macroscopic device sizes and kW energies to atomic domains and single-photon energies. The extreme versatility of the associated electromagnetic phenomena and applications, both classical and quantum, are therefore highly appealing to the rapidly evolving computing and communication realms, where innovations in both hardware and software are necessary to meet the growing speed and memory requirements. Development of all-optical components, photonic chips, interconnects, and processors will bring the speed of light, photon coherence properties, field confinement and enhancement, information-carrying capacity, and the broad spectrum of light into the high-performance computing, the internet of things, and industries related to cloud, fog, and recently edge computing. Conversely, owing to their extraordinary properties, 0D, 1D, and 2D materials are being explored as a physical basis for the next generation of logic components and processors. Carbon nanotubes, for example, have been recently used to create a new processor beyond proof of principle.
    [Show full text]
  • Advances in Photonic Devices Based on Optical Phase-Change Materials
    molecules Review Advances in Photonic Devices Based on Optical Phase-Change Materials Xiaoxiao Wang 1, Huixin Qi 1, Xiaoyong Hu 1,2,3,*, Zixuan Yu 1, Shaoqi Ding 1, Zhuochen Du 1 and Qihuang Gong 1,2,3 1 Collaborative Innovation Center of Quantum Matter & Frontiers Science Center for Nano-Optoelectronics, State Key Laboratory for Mesoscopic Physics & Department of Physics, Beijing Academy of Quantum Information Sciences, Peking University, Beijing 100871, China; [email protected] (X.W.); [email protected] (H.Q.); [email protected] (Z.Y.); [email protected] (S.D.); [email protected] (Z.D.); [email protected] (Q.G.) 2 Peking University Yangtze Delta Institute of Optoelectronics, Nantong 226010, China 3 Collaborative Innovation Center of Extreme Optics, Shanxi University, Taiyuan 030006, China * Correspondence: [email protected] Abstract: Phase-change materials (PCMs) are important photonic materials that have the advantages of a rapid and reversible phase change, a great difference in the optical properties between the crystalline and amorphous states, scalability, and nonvolatility. With the constant development in the PCM platform and integration of multiple material platforms, more and more reconfigurable photonic devices and their dynamic regulation have been theoretically proposed and experimentally demonstrated, showing the great potential of PCMs in integrated photonic chips. Here, we review the recent developments in PCMs and discuss their potential for photonic devices. A universal overview of the mechanism of the phase transition and models of PCMs is presented. PCMs have injected new life into on-chip photonic integrated circuits, which generally contain an optical switch, Citation: Wang, X.; Qi, H.; Hu, X.; an optical logical gate, and an optical modulator.
    [Show full text]
  • Optical Memristive Switches
    J Electroceram DOI 10.1007/s10832-017-0072-3 Optical memristive switches Ueli Koch1 & Claudia Hoessbacher1 & Alexandros Emboras1 & Juerg Leuthold1 Received: 18 September 2016 /Accepted: 6 February 2017 # The Author(s) 2017. This article is published with open access at Springerlink.com Abstract Optical memristive switches are particularly inter- signals. Normally, the operation speed is moderate and in esting for the use as latching optical switches, as a novel op- the MHz range. The application range includes usage as a tical memory or as a digital optical switch. The optical new kind of memory which can be electrically written and memristive effect has recently enabled a miniaturization of optically read, or usage as a latching switch that only needs optical devices far beyond of what seemed feasible. The to be triggered once and that can keep the state with little or no smallest optical – or plasmonic – switch has now atomic scale energy consumption. In addition, they represent a new logical andinfactisswitchedbymovingsingleatoms.Inthisreview, element that complements the toolbox of optical computing. we summarize the development of optical memristive The optical memristive effect has been discovered only switches on their path from the micro- to the atomic scale. recently [1]. It is of particular interest because of strong Three memristive effects that are important to the optical field electro-optical interaction with distinct transmission states are discussed in more detail. Among them are the phase tran- and because of low power consumption and scalability [8]. sition effect, the valency change effect and the electrochemical Such devices rely in part on exploiting the electrical metallization.
    [Show full text]
  • Optical Computing: a 60-Year Adventure Pierre Ambs
    Optical Computing: A 60-Year Adventure Pierre Ambs To cite this version: Pierre Ambs. Optical Computing: A 60-Year Adventure. Advances in Optical Technologies, 2010, 2010, pp.1-15. 10.1155/2010/372652. hal-00828108 HAL Id: hal-00828108 https://hal.archives-ouvertes.fr/hal-00828108 Submitted on 30 May 2013 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Hindawi Publishing Corporation Advances in Optical Technologies Volume 2010, Article ID 372652, 15 pages doi:10.1155/2010/372652 Research Article Optical Computing: A 60-Year Adventure Pierre Ambs Laboratoire Mod´elisation Intelligence Processus Syst`emes, Ecole Nationale Sup´erieure d’Ing´enieurs Sud Alsace, Universit´e de Haute Alsace, 12 rue des Fr`eres Lumi`ere, 68093 Mulhouse Cedex, France Correspondence should be addressed to Pierre Ambs, [email protected] Received 15 December 2009; Accepted 19 February 2010 Academic Editor: Peter V. Polyanskii Copyright © 2010 Pierre Ambs. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Optical computing is a very interesting 60-year old field of research.
    [Show full text]
  • Extending the Road Beyond CMOS
    By James A. Hutchby, George I. Bourianoff, Victor V. Zhirnov, and Joe E. Brewer he quickening pace of MOSFET technology scaling, as seen in the new 2001 International Technology Roadmap for Semiconductors [1], is ac- celerating the introduction of many new technologies to extend CMOS Tinto nanoscale MOSFET structures heretofore not thought possible. A cautious optimism is emerging that these new technologies may extend MOSFETs to the 22-nm node (9-nm physical gate length) by 2016 if not by the end of this decade. These new devices likely will feature several new materials cleverly incorporated into new nonbulk MOSFET structures. They will be ultra fast and dense with a voracious appetite for power. Intrinsic device speeds may be more than 1 THz and integration densities will exceed 1 billion transistors per cm2. Excessive power consumption, however, will demand judicious use of these high-performance devices only in those critical paths requiring their su- perior performance. Two or perhaps three other lower performance, more power-efficient MOSFETs will likely be used to perform less performance-criti- cal functions on the chip to manage the total power consumption. Beyond CMOS, several completely new approaches to information-process- ing and data-storage technologies and architectures are emerging to address the timeframe beyond the current roadmap. Rather than vying to “replace” CMOS, one or more of these embryonic paradigms, when combined with a CMOS plat- form, could extend microelectronics to new applications domains currently not accessible to CMOS. A successful new information-processing paradigm most likely will require a new platform technology embodying a fabric of intercon- nected primitive logic cells, perhaps in three dimensions.
    [Show full text]
  • Table of Contents
    Table of Contents Schedule-at-a-Glance . 2 FiO + LS Chairs’ Welcome Letters . 3 General Information . 5 Conference Materials Access to Technical Digest Papers . 7 FiO + LS Conference App . 7 Plenary Session/Visionary Speakers . 8 Science & Industry Showcase Theater Programming . 12 Networking Area Programming . 12 Participating Companies . 14 OSA Member Zone . 15 Special Events . 16 Awards, Honors and Special Recognitions FiO + LS Awards Ceremony & Reception . 19 OSA Awards and Honors . 19 2019 APS/Division of Laser Science Awards and Honors . 21 2019 OSA Foundation Fellowship, Scholarships and Special Recognitions . 21 2019 OSA Awards and Medals . 22 OSA Foundation FiO Grants, Prizes and Scholarships . 23 OSA Senior Members . 24 FiO + LS Committees . 27 Explanation of Session Codes . 28 FiO + LS Agenda of Sessions . 29 FiO + LS Abstracts . 34 Key to Authors and Presiders . 94 Program updates and changes may be found on the Conference Program Update Sheet distributed in the attendee registration bags, and check the Conference App for regular updates . OSA and APS/DLS thank the following sponsors for their generous support of this meeting: FiO + LS 2019 • 15–19 September 2019 1 Conference Schedule-at-a-Glance Note: Dates and times are subject to change. Check the conference app for regular updates. All times reflect EDT. Sunday Monday Tuesday Wednesday Thursday 15 September 16 September 17 September 18 September 19 September GENERAL Registration 07:00–17:00 07:00–17:00 07:30–18:00 07:30–17:30 07:30–11:00 Coffee Breaks 10:00–10:30 10:00–10:30 10:00–10:30 10:00–10:30 10:00–10:30 15:30–16:00 15:30–16:00 13:30–14:00 13:30–14:00 PROGRAMMING Technical Sessions 08:00–18:00 08:00–18:00 08:00–10:00 08:00–10:00 08:00–12:30 15:30–17:00 15:30–18:30 Visionary Speakers 09:15–10:00 09:15–10:00 09:15–10:00 09:15–10:00 LS Symposium on Undergraduate 12:00–18:00 Research Postdeadline Paper Sessions 17:15–18:15 SCIENCE & INDUSTRY SHOWCASE Science & Industry Showcase 10:00–15:30 10:00–15:30 See page 12 for complete schedule of programs .
    [Show full text]
  • Photonic Computing by Smit Patel What Is Photonic Computing?
    PHOTONIC COMPUTING BY SMIT PATEL WHAT IS PHOTONIC COMPUTING? • Optical Computing • Photonic computers perform its computations with photons or IR beams as opposed to electron-based computation which are seen in the traditional computers. • Two types: Pure Optical Computers & Electro-Optical Hybrid ELECTRO-OPTICAL HYBRID • Use optical fibers and electric parts to read and direct data from the processor Light pulses send information instead of voltage packets. • Processors change from binary code to light pulses using lasers. • Information is then detected and decoded electronically back into binary. OPTICAL TRANSISTORS • Transistors based off the Fabry-Perot Interferometer. • Constructive interference yields a high intensity (a 1 in binary) • Destructive interference yields an intensity close to zero (a 0 in binary) HOLOGRAPHIC MEMORY • A holographic memory can store data in the form of a hologram within a crystal. • A laser is split into a reference beam and a signal beam. • Signal beam goes through the logic gate and receives information • The two beams then meet up again and interference pattern creates a hologram in the crystal. HOLOGRAPHIC MEMORY OPTICAL FIBERS • Small in size • Low transmission losses • No interference from radio frequencies, electromagnetic components, or crosstalk • Safer • More secure • Environmental immunity OPTICAL FIBERS PROS • Small size • Increased speed • Low heating • Reconfigurable • Scalable for larger or small networks • More complex functions done faster Applications for Artificial Intelligence • Less power consumption (500 microwatts per interconnect length vs. 10 mW for electrical) LIMITS OF PHOTONIC COMPUTING • Optical fibers on a chip are wider than electrical traces. • Crystals need 1mm of length and are much larger than current transistors • Software needed to design and run the computers.
    [Show full text]
  • Non-Boolean Computing with Spintronic Devices
    Full text available at: http://dx.doi.org/10.1561/1000000046 Non-Boolean Computing with Spintronic Devices Kawsher A. Roxy University of South Florida [email protected] Sanjukta Bhanja University of South Florida [email protected] Boston — Delft Full text available at: http://dx.doi.org/10.1561/1000000046 Foundations and Trends R in Electronic Design Automation Published, sold and distributed by: now Publishers Inc. PO Box 1024 Hanover, MA 02339 United States Tel. +1-781-985-4510 www.nowpublishers.com [email protected] Outside North America: now Publishers Inc. PO Box 179 2600 AD Delft The Netherlands Tel. +31-6-51115274 The preferred citation for this publication is K. A. Roxy and S. Bhanja. Non-Boolean Computing with Spintronic Devices. Foundations and Trends R in Electronic Design Automation, vol. 12, no. 1, pp. 1–124, 2018. R This Foundations and Trends issue was typeset in LATEX using a class file designed by Neal Parikh. Printed on acid-free paper. ISBN: 978-1-68083-362-1 c 2018 K. A. Roxy and S. Bhanja All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, photocopying, recording or otherwise, without prior written permission of the publishers. Photocopying. In the USA: This journal is registered at the Copyright Clearance Cen- ter, Inc., 222 Rosewood Drive, Danvers, MA 01923. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by now Publishers Inc for users registered with the Copyright Clearance Center (CCC).
    [Show full text]
  • Chapter 8 Photonic Neuromorphic Signal Processing and Computing
    Chapter 8 Photonic Neuromorphic Signal Processing and Computing Alexander N. Tait, Mitchell A. Nahmias, Yue Tian, Bhavin J. Shastri and Paul R. Prucnal Abstract There has been a recent explosion of interest in spiking neural networks (SNNs), which code information as spikes or events in time. Spike encoding is widely accepted as the information medium underlying the brain, but it has also inspired a new generation of neuromorphic hardware. Although electronics can match biologi- cal time scales and exceed them, they eventually reach a bandwidth fan-in trade-off. An alternative platform is photonics, which could process highly interactive infor- mation at speeds that electronics could never reach. Correspondingly, processing techniques inspired by biology could compensate for many of the shortcomings that bar digital photonic computing from feasibility, including high defect rates and signal control problems. We summarize properties of photonic spike processing and ini- tial experiments with discrete components. A technique for mapping this paradigm to scalable, integrated laser devices is explored and simulated in small networks. This approach promises to wed the advantageous aspects of both photonic physics and unconventional computing systems. Further development could allow for fully scalable photonic networks that would open up a new domain of ultrafast, robust, and adaptive processing. Applications of this technology ranging from nanosecond response control systems to fast cognitive radio could potentially revitalize special- ized photonic computing. 8.1 Introduction The brain, unlike the von Neumann processors found in conventional computers, is very power efficient, extremely effective at certain computing tasks, and highly adaptable to novel situations and environments. While these favorable properties are often largely credited to the unique connectionism of biological nervous systems, it is A.
    [Show full text]
  • Optical Computing on Silicon-On-Insulator-Based Photonic Integrated Circuits
    Optical Computing on Silicon-on-Insulator-Based Photonic Integrated Circuits Zheng Zhao1, Zheng Wang2, Zhoufeng Ying1, Shounak Dhar1, Ray T. Chen1; 2, and David Z. Pan1 1Department of Electrical and Computer Engineering, University of Texas at Austin 2Department of Materials Science and Engineering, University of Texas at Austin [email protected], [email protected], [email protected], [email protected], [email protected] [email protected] ABSTRACT ing paradigms, logic functions and matrix multiplication, The advancement of photonic integrated circuits (PICs) both of which can be realized in SOI-based PIC. The ad- brings the possibility to accomplish on-chip optical in- vantages and limitations are also studied. The paper is terconnects and computations. Optical computing, as a concluded in Section4. promising alternative to traditional CMOS computing, has great potential advantages of ultra-high speed and low- 2. OPTICAL COMPUTING COMPONENTS power in information processing and communications. In In this section, we introduce the common SOI-based op- this paper, we survey the current research efforts on opti- tical components that demonstrate computing capabilities cal computing demonstrated on silicon-on-insulator-based as well as their working principles. PICs. The advantages, limitations, and possible research directions for further investigation are discussed. 2.1 Mach-Zehnder Interferometer The Mach-Zehnder interferometer (MZI) is a common 1. INTRODUCTION integrated photonic device used as electro-optic modula- tors and switches. In the 2 × 2 MZI in Figure 1a, the Optical computing is to use optical systems to perform lights from input ports a, b go into a splitter, traveling numerical computations [1].
    [Show full text]
  • Myths and Truths About Optical Phase Change Materials: a Perspective
    Myths and truths about optical phase change materials: A perspective Cite as: Appl. Phys. Lett. 118, 210501 (2021); https://doi.org/10.1063/5.0054114 Submitted: 14 April 2021 . Accepted: 11 May 2021 . Published Online: 26 May 2021 Yifei Zhang, Carlos Ríos, Mikhail Y. Shalaginov, Mo Li, Arka Majumdar, Tian Gu, and Juejun Hu ARTICLES YOU MAY BE INTERESTED IN Perspective on the future of silicon photonics and electronics Applied Physics Letters 118, 220501 (2021); https://doi.org/10.1063/5.0050117 Enhanced light-matter interactions at photonic magic-angle topological transitions Applied Physics Letters 118, 211101 (2021); https://doi.org/10.1063/5.0052580 Ultrawide bandgap semiconductors Applied Physics Letters 118, 200401 (2021); https://doi.org/10.1063/5.0055292 Appl. Phys. Lett. 118, 210501 (2021); https://doi.org/10.1063/5.0054114 118, 210501 © 2021 Author(s). Applied Physics Letters PERSPECTIVE scitation.org/journal/apl Myths and truths about optical phase change materials: A perspective Cite as: Appl. Phys. Lett. 118, 210501 (2021); doi: 10.1063/5.0054114 Submitted: 14 April 2021 . Accepted: 11 May 2021 . Published Online: 26 May 2021 Yifei Zhang,1 Carlos Rıos,1 Mikhail Y. Shalaginov,1 Mo Li,2,3 Arka Majumdar,2,3 Tian Gu,1,4,a) and Juejun Hu1,4,a) AFFILIATIONS 1Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA 2Department of Electrical and Computer Engineering, University of Washington, Seattle, Washington 98195, USA 3Department of Physics, University of Washington, Seattle, Washington 98195, USA 4Materials Research Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA a)Authors to whom correspondence should be addressed: [email protected] and [email protected] ABSTRACT Uniquely furnishing giant and nonvolatile modulation of optical properties and chalcogenide phase change materials (PCMs) have emerged as a promising material to transform integrated photonics and free-space optics alike.
    [Show full text]