Diffraction casting

TL;DR

This paper introduces diffraction casting, a novel optical computing architecture using diffractive neural networks for scalable, high-speed parallel logic operations that enable flexible, end-to-end all-optical processing without encoding complexities.

Contribution

The paper proposes a new optical computation architecture called diffraction casting that utilizes diffractive neural networks for flexible, scalable SIMD logic operations with simplified input/output handling.

Findings

Successfully demonstrated all 16 logic operations on 256-bit inputs.

Highlights high scalability and integration potential of the architecture.

Enables flexible logic operation changes via illumination pattern adjustments.

Abstract

Optical computing is considered a promising solution for the growing demand for parallel computing in various cutting-edge fields, requiring high integration and high speed computational capacity. In this paper, we propose a novel optical computation architecture called diffraction casting (DC) for flexible and scalable parallel logic operations. In DC, a diffractive neural network (DNN) is designed for single instruction, multiple data (SIMD) operations. This approach allows for the alteration of logic operations simply by changing the illumination patterns. Furthermore, it eliminates the need for encoding and decoding the input and output, respectively, by introducing a buffer around the input area, facilitating end-to-end all-optical computing. We numerically demonstrate DC by performing all 16 logic operations on two arbitrary 256 bits parallel binary inputs. Additionally, we showcase several distinctive attributes inherent in DC, such as the benefit of cohesively designing the diffractive elements for SIMD logic operations, assuring high scalability and integration capability. Our study offers a novel design architecture for optical computers and paves the way for a next-generation optical computing paradigm.