English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

An emerging generative artificial intelligence (AI) based on neural networksstarts to grow in popularity with a revolutionizing capability of creating newand original content. As giant generative models with millions to billions ofparameters are developed, trained and maintained, a massive andenergy-efficient computing power is highly required. However, conventionaldigital computers are struggling to keep up with the pace of the generativemodel improvements. In this paper, we propose and demonstrate high-throughputoptical neural networks based on temporal computing. The core weightedsummation operation is realized with the use of high-speed electro-opticmodulation and low-speed balanced photodetection. The input data and weight areencoded in a time sequence separately and loaded on an optical signal via twoelectro-optic modulators sequentially. By precisely controlling thesynchronization time of the data and weight loading, the matrix multiplicationis performed. Followed by a balanced photodetector, the summation is conducted,thanks to the electron accumulation of the inherent electronic integratorcircuit of the low-speed photodetector. Thus, the linear weighted summationoperation is implemented based on temporal computing in the optical domain.With the proposed optical linear weighted summation, a fully-connected neuralnetwork and convolutional neural network are realized. Thanks to the high-speedfeature of temporal computing, a high data throughput of the optical neuralnetwork is experimentally demonstrated, and the weighting coefficients can bespecified on demand, which enables a strong programmability of the opticalneural network. By leveraging wavelength multiplexing technology, a scalableoptical neural network could be created with a massive computing power andstrong reconfigurability, which holds great potential for future giant AIapplications.

High-throughput optical neural networks based on temporal computing