Photonic Neuromorphic Accelerator for Convolutional Neural Networks based on an Integrated Reconfigurable Mesh

TL;DR

This paper introduces a passive photonic neuromorphic accelerator using a reconfigurable silicon photonic mesh for efficient convolutional neural network processing, achieving high accuracy with reduced power consumption.

Contribution

It presents a novel passive photonic integrated neuromorphic accelerator leveraging optical spectrum slicing and a reconfigurable mesh, enabling efficient CNN modules with fewer passive nodes.

Findings

Achieved 98.6% accuracy on MNIST with 7 passive nodes.

Reduced power consumption by at least 26% compared to digital CNNs.

Experimental validation confirmed 97.7% accuracy with only 3 passive nodes.

Abstract

In this work, we present and experimentally validate a passive photonic-integrated neuromorphic accelerator that uses a hardware-friendly optical spectrum slicing technique through a reconfigurable silicon photonic mesh. The proposed scheme acts as an analogue convolutional engine, enabling information preprocessing in the optical domain, dimensionality reduction and extraction of spatio-temporal features. Numerical results demonstrate that utilizing only 7 passive photonic nodes, critical modules of a digital convolutional neural network can be replaced. As a result, a 98.6% accuracy on the MNIST dataset was achieved, with a power consumption reduction of at least 26% compared to digital CNNs. Experimental results confirm these findings, achieving 97.7% accuracy with only 3 passive nodes.