面向下一代人工智能计算的资源高效光子网络。

Resource-efficient photonic networks for next-generation AI computing.

作者信息

Oguz Ilker, Yildirim Mustafa, Hsieh Jih-Liang, Dinc Niyazi Ulas, Moser Christophe, Psaltis Demetri

机构信息

EPFL, Institute of Electrical and Micro Engineering, Lausanne, Switzerland.

出版信息

Light Sci Appl. 2025 Jan 4;14(1):34. doi: 10.1038/s41377-024-01717-6.

DOI:10.1038/s41377-024-01717-6

PMID:39753539

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC11698740/

Abstract

Current trends in artificial intelligence toward larger models demand a rethinking of both hardware and algorithms. Photonics-based systems offer high-speed, energy-efficient computing units, provided algorithms are designed to exploit photonics' unique strengths. The recent implementation of cellular automata in photonics demonstrates how a few local interactions can achieve high throughput and precision.

摘要

当前人工智能朝着更大模型发展的趋势要求重新思考硬件和算法。基于光子学的系统提供了高速、节能的计算单元，前提是算法设计要利用光子学的独特优势。光子学中细胞自动机的最新实现展示了一些局部相互作用如何实现高吞吐量和高精度。