An inversely designed mode-multiplexed photonic integrated vector dot-product core

Photonic computing has the potential of harnessing the full degrees of freedom (DOFs) of the light field, including wavelength, spatial mode, spatial location, phase quadrature, and polarization, to achieve higher level of computation parallelization and scalability than digital electronic processors. While multiplexing using wavelength and other DOFs can be readily integrated on silicon photonics platforms with compact footprints, conventional mode-division multiplexed (MDM) photonic designs occupy areas exceeding tens to hundreds of microns for a few spatial modes, significantly limiting their scalability. Here we utilize inverse design to demonstrate an ultracompact photonic computing core that calculates vector dot-products based on MDM coherent mixing within a nominal footprint of 5 um x 3 um. Our dot-product core integrates the functionalities of 2 mode multiplexers and 1 multi-mode coherent mixers, all within the footprint, and could be applied to various computation and computer vision tasks, with high computing throughput density. We experimentally demonstrate computing examples on the fabricated core, including complex number multiplication and motion estimation using optical flow.

A physical neural network training approach toward multi-plane light conversion design

Multi-plane light converter (MPLC) designs supporting hundreds of modes are attractive in high-throughput optical communications. These photonic structures typically comprise >10 phase masks in free space, with millions of independent design parameters. Conventional MPLC design using wavefront matching updates one mask at a time while fixing the rest. Here we construct a physical neural network (PNN) to model the light propagation and phase modulation in MPLC, providing access to the entire parameter set for optimization, including not only profiles of the phase masks and the distances between them. PNN training supports flexible optimization sequences and is a superset of existing MPLC design methods. In addition, our method allows tuning of hyperparameters of PNN training such as learning rate and batch size. Because PNN-based MPLC is found to be insensitive to the number of input and target modes in each training step, we have demonstrated a high-order MPLC design (45 modes) using mini batches that fit into the available computing resources.

Training of mixed-signal optical convolutional neural network with reduced quantization level

Mixed-signal artificial neural networks (ANNs) that employ analog matrix-multiplication accelerators can achieve higher speed and improved power efficiency. Though analog computing is known to be susceptible to noise and device imperfections, various analog computing paradigms have been considered as promising solutions to address the growing computing demand in machine learning applications, thanks to the robustness of ANNs. This robustness has been explored in low-precision, fixed-point ANN models, which have proven successful on compressing ANN model size on digital computers. However, these promising results and network training algorithms cannot be easily migrated to analog accelerators. The reason is that digital computers typically carry intermediate results with higher bit width, though the inputs and weights of each ANN layers are of low bit width; while the analog intermediate results have low precision, analogous to digital signals with a reduced quantization level. Here we report a training method for mixed-signal ANN with two types of errors in its analog signals, random noise, and deterministic errors (distortions). The results showed that mixed-signal ANNs trained with our proposed method can achieve an equivalent classification accuracy with noise level up to 50% of the ideal quantization step size. We have demonstrated this training method on a mixed-signal optical convolutional neural network based on diffractive optics.