Double-layer networks with holographic optical switches

Deng, Jiun-Shiou; Lu, Ming–Feng; Huang, Yang‐Tung

doi:10.1364/ao.43.001342

Cited by 5 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, we theoretically and experimentally demonstrate an easily scalable MZI network topology to carry out calculations for arbitrary real-and nonnegative-valued matrices for PNNs, similar to dilated double-layer networks (SDDLNs). [28][29][30] By delicately designed connection topology, SDDLNs are demonstrated to have two orders of magnitude lower optical path loss and a twofold reduction in the number of tunable phase shifters compared with the universal matrix mesh in the implementation of the largescale matrix. We fabricated a reconfigurable 8 × 8 MZI-based array and experimentally verified the proposed structure with the implementation of a 2 × 4 nonnegative-valued matrix and a 2 × 2 real-valued matrix.…”

Section: Introductionmentioning

confidence: 99%

Easily Scalable Photonic Tensor Core Based on Tunable Units with Single Internal Phase Shifters

Huang

Yue

et al. 2023

Laser & Photonics Reviews

View full text Add to dashboard Cite

Photonic neural networks (PNNs) show tremendous potential for artificial intelligence applications due to their higher computational rates than their traditional electronic counterpart. However, the scale‐up of PNN relies on the number of cascaded computing units, which is limited by the accumulated transmission attenuation. Here, a topology of PNN with Mach–Zehnder interferometers based on a single‐tuned phase shifter that implements arbitrary nonnegative or real‐valued matrices for vector‐matrix multiplication is proposed. Compared with the universal matrix mesh, the new configuration exhibits two orders of magnitude lower optical path loss and a twofold reduction in the number of the tunable phase shifter. An 8 × 8 reconfigurable chip is designed and fabricated, and it is experimentally verified that the 2 × 4 nonnegative‐valued matrix and the 2 × 2 real‐valued matrix are implemented in the proposed topology. Higher than 85% inference accuracies are obtained in the Modified National Institute of Standards and Technology handwritten digit recognition tasks with these matrices in the PNNs. Therefore, with much lower optical path loss and comparable computing accuracy, the proposed PNN configuration can be easily scaled up to tackle higher dimensional matrix multiplication, which is highly desired in tasks like voice and image recognition.

show abstract

Section: Introductionmentioning

confidence: 99%