The physics of optical computing

Abstract

There has been a resurgence of interest in optical computing over the past decade, both in academia and in industry, with much of the excitement centered around special-purpose optical computers for neural-network processing. Optical computing has been a topic of periodic study for over 50 years, including for neural networks three decades ago, and a wide variety of optical-computing schemes and architectures have been proposed. In this paper we provide a systematic explanation of why and how optics might be able to give speed or energy-efficiency benefits over electronics for computing, enumerating 11 features of optics that can be harnessed when designing an optical computer. One often-mentioned motivation for optical computing -- that the speed of light $c$ is fast -- is not a key differentiating physical property of optics for computing; understanding where an advantage could come from is more subtle. We discuss how gaining an advantage over state-of-the-art electronic processors will likely only be achievable by careful design that harnesses more than one of the 11 features, while avoiding a number of pitfalls that we describe.

arXiv comment31 pages; 11 figures

References in corpus (38)

• Deep Learning with Coherent Nanophotonic Circuits
• Parallel convolution processing using an integrated photonic tensor core
• Photonics for artificial intelligence and neuromorphic computing
• 11 TeraFLOPs per second photonic convolutional accelerator for deep learning optical neural networks
• Microresonator solitons for massively parallel coherent optical communications
• Universal Linear Optics
• Deep physical neural networks enabled by a backpropagation algorithm for arbitrary physical systems
• Single-chip photonic deep neural network for instantaneous image classification
• Attojoule Optoelectronics for Low-Energy Information Processing and Communications: a Tutorial Review
• Large-scale neuromorphic optoelectronic computing with a reconfigurable diffractive processing unit
• Synthetic three-dimensional atomic structures assembled atom by atom
• Network of Time-Multiplexed Optical Parametric Oscillators as a Coherent Ising Machine
• 30 years of squeezed light generation
• Reprogrammable Electro-Optic Nonlinear Activation Functions for Optical Neural Networks
• Ultra-low loss integrated visible photonics using thin-film lithium niobate
• An optical neural network using less than 1 photon per multiplication
• Image sensing with multilayer, nonlinear optical neural networks
• Superconducting optoelectronic circuits for neuromorphic computing
• Delocalized Photonic Deep Learning on the Internet's Edge
• Scalable Optical Learning Operator
• Prospects and applications of photonic neural networks
• Reduced Material Loss in Thin-film Lithium Niobate Waveguides
• Femtojoule, femtosecond all-optical switching in lithium niobate nanophotonics
• Three dimensional waveguide-interconnects for scalable integration of photonic neural networks
• Random Projections through multiple optical scattering: Approximating kernels at the speed of light
• A full degree-of-freedom photonic crystal spatial light modulator
• An Optical Frontend for a Convolutional Neural Network
• Design and analysis of photonic crystal coupled cavity arrays for quantum simulation
• Asymptotically Fault-Tolerant Programmable Photonics
• Understanding and mitigating noise in trained deep neural networks
• Quantifying power use in silicon photonic neural networks
• Remnants of semiclassical bistability in the few-photon regime of cavity QED
• Nonlinear interferometry approach to photonic sequential logic
• Macroscopic electron quantum coherence in a solid-state circuit
• Optical Neural Networks: The 3D connection
• Thermalization of strongly interacting bosons after spontaneous emissions in optical lattices
• Photonic unsupervised learning processor for secure and high-throughput optical fiber communication
• On the physical (im)possibility of lightsabers

Cited by in corpus (37)

• Intrinsic Optical Bistability of Photon Avalanching Nanocrystals
• Programmable on-chip nonlinear photonics
• Low threshold integrated optical parametric oscillator with a compact Bragg resonator
• Symmetric silicon microring resonator optical crossbar array for accelerated inference and training in deep learning
• Programmable metasurfaces for future photonic artificial intelligence
• Transferable polychromatic optical encoder for neural networks
• Waveguide-multiplexed photonic matrix-vector multiplication processor using multiport photodetectors
• Asymmetrical estimator for training encapsulated deep photonic neural networks
• XLuminA: An Auto-differentiating Discovery Framework for Super-Resolution Microscopy
• Programmable 200 GOPS Hopfield-inspired photonic Ising machine
• Dawn and fall of non-Gaussianity in the quantum parametric oscillator
• Meta-operators for all-optical image processing
• Multi-task Photonic Reservoir Computing: Wavelength Division Multiplexing for Parallel Computing with a Silicon Microring Resonator
• Arbitrary control over multimode wave propagation for machine learning
• Large-scale free-space photonic circuits in two dimensions
• Fault-Tolerant Neural Networks from Biological Error Correction Codes
• Programmable Space-Frequency Linear Transformations in Photonic Interlacing Architectures
• An All-Optical General-Purpose CPU and Optical Computer Architecture
• Coherence Awareness in Diffractive Neural Networks
• Photonic systolic array for all-optical matrix-matrix multiplication
• Quantum optical classifier with superexponential speedup
• Mechanical in-sensor computing: a programmable meta-sensor for structural damage classification without external electronic power
• Nonlinear unitary circuits for photonic neural networks
• Passive All-Optical Nonlinear Neuron Activation via PPLN Nanophotonic Waveguides
• Quantum-secure multiparty deep learning
• Spatial Optical Simulator for Classical Statistical Models
• Dense Associative Memory in a Nonlinear Optical Hopfield Neural Network
• Optical Algorithm for Derivative of Real-Valued Functions
• Neural networks with quantum states of light
• Shannon Entropy Helps Optimize the Performance of a Frequency-Multiplexed Extreme Learning Machine
• Polaritonic Machine Learning for Graph-based Data Analysis
• Scaling Analog Photonic Accelerators for Byte-Size, Integer General Matrix Multiply (GEMM) Kernels
• Integrated Photonic Devices in Thin-Film Barium Titanate: Opportunities and Challenges
• Implementation of transformer-based LLMs with large-scale optoelectronic neurons on a CMOS compatible platform
• Electrically driven plasmon-polaritonic bistability in Dirac electron tunneling transistors
• A ghost mechanism: An analytical model of abrupt learning in recurrent networks
• Switching, Amplifying, and Chirping Diode Lasers with Current Pulses for High Bandwidth Quantum Technologies