Parallel photonic acceleration processor for matrix-matrix multiplication

We propose and experimentally demonstrate a highly par-allel photonic acceleration processor based on a wave-length division multiplexing (WDM) system and a non -coherent Mach-Zehnder interferometer (MZI) array for matrix-matrix multiplication. The dimensional expansion is achieved by WDM devices, which play a crucial role in realizing matrix-matrix multiplication together with the broadband characteristics of an MZI. We implemented a 2 x 2 arbitrary nonnegative valued matrix using a reconfig-urable 8 x 8 MZI array structure. Through experimentation, we verified that this structure could achieve 90.5% infer-ence accuracy in a classification task for the Modified National Institute of Standards and Technology (MNIST) handwritten dataset. This provides a new effective solution for large-scale integrated optical computing systems based on convolution acceleration processors.& COPY; 2023 Optica Pub-lishing Group

Automated summary

We propose and demonstrate a highly parallel photonic acceleration processor based on a wavelength division multiplexing (WDM) system and a non-coherent Mach-Zehnder interferometer (MZI) array. The use of WDM devices enables dimensional expansion and, combined with the broadband characteristics of an MZI, allows for matrix-matrix multiplication. Through experiments, we achieve an inference accuracy of 90.5% in a classification task using a reconfigurable 8x8 MZI array structure for a 2x2 non-negative valued matrix. This research provides a new effective solution for large-scale integrated optical computing systems based on convolution acceleration processors.