Optical Neural Networks: The 3D connection

TL;DR

This paper proposes a novel 3D printing-based approach to photonic neural networks, enabling scalable dense connectivity that surpasses traditional 2D CMOS and photonics architectures.

Contribution

It introduces a new 3D integration strategy for photonic neural networks, leveraging additive fabrication to enhance connectivity and scalability.

Findings

Demonstrated the potential of 3D printing for neural network integration

Showed that 3D architecture can significantly improve connectivity density

Reviewed recent advances towards 3D photonic neural network architectures

Abstract

We motivate a new canonical strategy for integrating photonic neural networks (NNs) by leveraging 3D printing. Our believe is that a NN's parallel and dense connectivity is not scalable without 3D integration. 3D additive fabrication complemented with photonic signal transduction can dramatically augment the current capabilities of 2D CMOS and integrated photonics. Here we review some of our recent advances made towards such a breakthrough architecture.