Photonic Convolution Neural Network Based on Interleaved Time-Wavelength Modulation

TL;DR

This paper introduces a novel integrated photonic CNN utilizing interleaved time-wavelength modulation to perform convolution operations efficiently, aiming to reduce computational complexity and resource usage in high-definition video processing.

Contribution

It proposes a new photonic CNN architecture based on double correlation and micro-ring manipulation, replacing traditional optical delay lines, and demonstrates a parallel tensor processing unit for enhanced performance.

Findings

Achieved 85.5% accuracy on MNIST with photonic CNN

Analyzed error sources related to micro-ring parameters and operation rates

Proposed a 4x4 mesh tensor processing unit with 1.2 TOPS capability

Abstract

Convolution neural network (CNN), as one of the most powerful and popular technologies, has achieved remarkable progress for image and video classification since its invention in 1989. However, with the high definition video-data explosion, convolution layers in the CNN architecture will occupy a great amount of computing time and memory resources due to high computation complexity of matrix multiply accumulate operation. In this paper, a novel integrated photonic CNN is proposed based on double correlation operations through interleaved time-wavelength modulation. Micro-ring based multi-wavelength manipulation and single dispersion medium are utilized to realize convolution operation and replace the conventional optical delay lines. 200 images are tested in MNIST datasets with accuracy of 85.5% in our photonic CNN versus 86.5% in 64-bit computer.We also analyze the computing error of photonic CNN caused by various micro-ring parameters, operation baud rates and the characteristics of micro-ring weighting bank. Furthermore, a tensor processing unit based on 4x4 mesh with 1.2 TOPS (operation per second when 100% utilization) computing capability at 20G baud rate is proposed and analyzed to form a paralleled photonic CNN.