Quantifying power use in silicon photonic neural networks

Abstract

Due to challenging efficiency limits facing conventional and unconventional electronic architectures, information processors based on photonics have attracted renewed interest. Research communities have yet to settle on definitive techniques to describe the performance of this class of information processors. Photonic systems are different from electronic ones, so the existing concepts of computer performance measurement cannot necessarily apply. In this manuscript, we attempt to quantify the power use of photonic neural networks with state-of-the-art and future hardware. We derive scaling laws, physical limits, and new platform performance metrics. We find that overall performance is regime-like, which means that energy efficiency characteristics of a photonic processor can be completely described by no less than seven performance numbers. The introduction of these analytical strategies provides a much needed foundation for quantitative roadmapping and commercial value assignment for silicon photonic neural networks.

arXiv comment9 figures

References in corpus (23)

• Deep Learning with Coherent Nanophotonic Circuits
• Neuromorphic Silicon Photonic Networks
• A one femtojoule athermal silicon modulator
• Large-Scale Optical Neural Networks based on Photoelectric Multiplication
• Reprogrammable Electro-Optic Nonlinear Activation Functions for Optical Neural Networks
• A silicon photonic modulator neuron
• Digital Electronics and Analog Photonics for Convolutional Neural Networks (DEAP-CNNs)
• All-optical mode unscrambling on a silicon photonic chip
• Superconducting optoelectronic circuits for neuromorphic computing
• Proposal for an All-Spin Artificial Neural Network: Emulating Neural and Synaptic Functionalities Through Domain Wall Motion in Ferromagnets
• A BaTiO3-Based Electro-Optic Pockels Modulator Monolithically Integrated on an Advanced Silicon Photonics Platform
• WRPN: Wide Reduced-Precision Networks
• MEMS for Photonic Integrated Circuits
• Wavelength tuning and stabilization of microring-based filters using silicon in-resonator photoconductive heaters
• Neuromorphic photonics with electro-absorption modulators
• Is a 4-bit synaptic weight resolution enough? - Constraints on enabling spike-timing dependent plasticity in neuromorphic hardware
• Post-Fabrication Trimming of Silicon Photonic Ring Resonators at Wafer-Scale
• Noise Analysis of Photonic Modulator Neurons
• MorphoNoC: Exploring the Design Space of a Configurable Hybrid NoC using Nanophotonics
• In-Datacenter Performance Analysis of a Tensor Processing Unit
• Demonstration of multivariate photonics: blind dimensionality reduction with analog integrated photonics
• Accelerated Analog Neuromorphic Computing
• Single photonic perceptron based on a soliton crystal Kerr microcomb for high-speed, scalable, optical neural networks

Cited by in corpus (10)

• The physics of optical computing
• Silicon Photonic Architecture for Training Deep Neural Networks with Direct Feedback Alignment
• Broadband physical layer cognitive radio with an integrated photonic processor for blind source separation
• Integrated Photonic Tensor Processing Unit for a Matrix Multiply: a Review
• Quantum-limited stochastic optical neural networks operating at a few quanta per activation
• Real-Time Blind Photonic Interference Cancellation for mmWave MIMO
• Lower-depth programmable linear optical processors
• AnalogVNN: A fully modular framework for modeling and optimizing photonic neural networks
• The Impact of Feature Representation on the Accuracy of Photonic Neural Networks
• Online training and pruning of multi-wavelength photonic neural networks