Photonic systolic array for all-optical matrix-matrix multiplication

Abstract

Systolic arrays have proven to be highly efficient for parallelized matrix-matrix multiplication (MMM), utilizing synchronized, heartbeat-like data flows across an array of processing elements. While optical structures such as waveguide crossbar arrays and Mach-Zehnder interferometer-based meshes serve as photonic equivalents to the systolic arrays, the disparity between the two input matrices for multiplication -- one using optical signals and the other with system-defined parameters -- gives rise to a bottleneck in modern machine-learning tasks, such as evaluating attention scores in large language models. Here, we propose a photonic systolic array that performs MMM entirely with optical signals, utilizing homodyne detection at each array cell. Adjoint-based design of compact on-chip freeform optical modules enables precise control of light flow without bulky waveguide coupling schemes. The operation of a $4\times 4$ photonic systolic array is numerically verified, achieving a theoretical computation density of $6.2\mathrm{PMACs}/{\mathrm{mm}}^2/\mathrm{s}$. This design marks a significant step toward practical photonic computing hardware for modern AI workloads.

Main 8 pages, 4 figures Supplementary Note S1-2, Supplementary Figures S1-9

References in corpus (15)

• Deep Learning with Coherent Nanophotonic Circuits
• All-Optical Machine Learning Using Diffractive Deep Neural Networks
• Parallel convolution processing using an integrated photonic tensor core
• Photonics for artificial intelligence and neuromorphic computing
• Universal Linear Optics
• Single-chip photonic deep neural network for instantaneous image classification
• Training of photonic neural networks through in situ backpropagation
• The physics of optical computing
• Large-Scale Optical Neural Networks based on Photoelectric Multiplication
• Adjoint method and inverse design for nonlinear nanophotonic devices
• An optical neural network using less than 1 photon per multiplication
• Deep Learning with Coherent VCSEL Neural Networks
• High-order tensor flow processing using integrated photonic circuits
• Inverse-designed low-index-contrast structures on silicon photonics platform for vector-matrix multiplication
• A single inverse-designed photonic structure that performs parallel computing