Digital Electronics and Analog Photonics for Convolutional Neural Networks (DEAP-CNNs)

Abstract

Convolutional Neural Networks (CNNs) are powerful and highly ubiquitous tools for extracting features from large datasets for applications such as computer vision and natural language processing. However, a convolution is a computationally expensive operation in digital electronics. In contrast, neuromorphic photonic systems, which have experienced a recent surge of interest over the last few years, propose higher bandwidth and energy efficiencies for neural network training and inference. Neuromorphic photonics exploits the advantages of optical electronics, including the ease of analog processing, and busing multiple signals on a single waveguide at the speed of light. Here, we propose a Digital Electronic and Analog Photonic (DEAP) CNN hardware architecture that has potential to be 2.8 to 14 times faster while maintaining the same power usage of current state-of-the-art GPUs.

arXiv comment12 pages, 9 figures, 3 tables

References in corpus (11)

• Deep Learning with Coherent Nanophotonic Circuits
• cuDNN: Efficient Primitives for Deep Learning
• An Introduction to Convolutional Neural Networks
• Speeding up Convolutional Neural Networks with Low Rank Expansions
• Training of photonic neural networks through in situ backpropagation
• A silicon photonic modulator neuron
• Training Deep Neural Networks with 8-bit Floating Point Numbers
• Wavelength tuning and stabilization of microring-based filters using silicon in-resonator photoconductive heaters
• PCNNA: A Photonic Convolutional Neural Network Accelerator
• On-Chip Optical Convolutional Neural Networks
• Noise Analysis of Photonic Modulator Neurons

Cited by in corpus (19)

• Parallel convolution processing using an integrated photonic tensor core
• Photonics for artificial intelligence and neuromorphic computing
• Programmable Phase-change Metasurfaces on Waveguides for Multimode Photonic Convolutional Neural Network
• Prospects and applications of photonic neural networks
• High-order tensor flow processing using integrated photonic circuits
• Misalignment Resilient Diffractive Optical Networks
• Free-space optical neural network based on thermal atomic nonlinearity
• Silicon Photonic Architecture for Training Deep Neural Networks with Direct Feedback Alignment
• Scaling Up Silicon Photonic-based Accelerators: Challenges and Opportunities
• Noise-enhanced spatial-photonic Ising machine
• An Electro-Photonic System for Accelerating Deep Neural Networks
• Quantifying power use in silicon photonic neural networks
• Photonic Computing to Accelerate Data Processing in Wireless Communications
• Integration of Programmable Diffraction with Digital Neural Networks
• Channel Tiling for Improved Performance and Accuracy of Optical Neural Network Accelerators
• Performance evaluation of an integrated photonic convolutional neural network based on delay buffering and wavelength division multiplexing
• Prediction of Ultraslow Magnetic Solitons via Plasmon-induced Transparency by Artificial Neural Networks
• Scaling Analog Photonic Accelerators for Byte-Size, Integer General Matrix Multiply (GEMM) Kernels
• Optical Proof of Work