A scalable and programmable optical neural network in a time-synthetic dimension

TL;DR

This paper demonstrates a highly scalable, all-optical neural network operating in a time-synthetic dimension, overcoming traditional spatial limitations and enabling efficient, programmable AI computations with reduced errors.

Contribution

It introduces the first experimental all-optical neural network in a time-synthetic dimension, utilizing time-reflection and refraction for scalable, error-free computation, and an in-situ training framework.

Findings

Achieved a gate count surpassing state-of-the-art photonic processors.

Demonstrated error-free computation via causality constraints.

Implemented in-situ training for adaptive performance.

Abstract

Programmable optical neural networks (ONNs) can offer high-throughput and energy-efficient solutions for accelerating artificial intelligence (AI) computing. However, existing ONN architectures, typically based on cascaded unitary transformations such as Mach-Zehnder interferometer meshes, face inherent scalability limitations due to spatial encoding, which causes optical components and system complexity to scale quadratically with network size. A promising solution to this challenge is the use of synthetic dimensions to enhance scalability, though experimental demonstration has remained scarce. Here, we present the first experimental demonstration of an all-optical, highly scalable, programmable ONN operating in a time-synthetic dimension. By implementing a time-cycle computation paradigm analogous to gate cycling in conventional spatial photonic circuits, our approach achieves a gate count surpassing that of state-of-the-art programmable photonic processors. Unlike conventional ONN architectures that rely on real-space wave interferences, our framework exploits time-reflection and time-refraction to perform computations, fundamentally eliminating backscattering errors through causality constraints. To bridge the gap between simulation and reality, we introduce an in-situ training framework that dynamically adapts to experimental errors, achieving performance exceeding traditional in silico learning paradigms. Our synthetic-dimension-based approach provides a compact, scalable, backscattering-free, and programmable neuromorphic computing architecture, advancing the potential for next-generation photonic AI systems.