A semiconductor device capable of retaining a signal sensed by a sensor element is provided. The semiconductor device includes a sensor element, a first transistor, a second transistor, and a third transistor. One electrode of the sensor element is electrically connected to a first gate. The first gate is electrically connected to one of a source and a drain of the third transistor. One of a source and a drain of the first transistor is electrically connected to a gate of the second transistor. A semiconductor layer includes a metal oxide.

A method for assembling a photonic computing system includes attaching a photonic source to a support structure, and attaching a photonic integrated circuit to the support structure. The photonic source includes a first laser die on a substrate configured to provide a first optical beam, and a second laser die on the substrate configured to provide a second optical beam. The photonic integrated circuit includes a first waveguide and a first coupler coupled to the first waveguide, and a second waveguide and a second coupler coupled to the second waveguide. The method includes attaching a plurality of beam-shaping optical elements to the support structure, the substrate, or the photonic integrated circuit, in which the attaching includes aligning a first beam-shaping optical element during attachment so that the first optical beam is coupled to the first coupler, and aligning a second beam-shaping optical element during attachment so that the second optical beam is coupled to the second coupler.

Utilizing the principles of wavelength-dependent evanescent wave coupling in closely-spaced optical waveguides, along with optical resonators, a method for creating a neural network out of entirely electro-optical components is discussed. Optical resonators, which can store energy as standing waves or whispering gallery modes, act as neurons. Waveguides integrated onto a chip act as dendrites or connectomes, with coupling between them simulating the analog exchange of signals in brains. Additional electro-optic controls can be utilized, such as conductive plates utilizing the electro-optic effect to change the refractive indices of the optics and coupling coefficients based on electrical signals from outside stimuli.

An optical neural network having at least one layer including: an optical transmission element arranged such that the signal of each node passes through the optical transmission element in both forward and backpropagation; wherein the optical transmission element comprises a saturable optical absorption material or a saturable optical gain material, having a saturation threshold-power; wherein optical signals propagating in a forward direction have a power below the saturation threshold-power at least some of the time, such that transmission of the optical signal through the optical transmission element in a forward direction is nonlinear; and wherein optical signals propagating in a backward direction have a power below a second threshold-power, lower than the saturation threshold-power, and transmission of the optical signal in a backward direction through the optical transmission element is approximately linear.

Photonic units for vector-vector multiplication, matrix-vector multiplication, matrix-matrix multiplication, batch matrix-matrix multiplication, and tensor-tensor multiplication are described. Multiplications are through coherent mixing and square-law detection. There are many dimensions—wavelength, vector mode, quadrature, and three dimensions of space—that can be used to construct photonic accelerators. The encoded input vector or input matrix is fanned out into a desired number of copies and mixed with the corresponding encoded local oscillators containing the weight vectors comprising the weight matrix. Any subset of two (three) dimensions can be used to construct photonic accelerators for matrix-vector (matrix-matrix) multiplications. Multiple dimensions can be combined into a hyperdimension to increase the scalability. Each dimension, each non-overlapping subset of a dimension, or each non-overlapping subset of a hyperdimension, can be used independently to construct a photonic tensor accelerator (PTA) for batch matrix multiplication, or tensor multiplication operations.

An optical diffractive processing unit includes input nodes, output nodes; and neurons. The neurons are connected to the input nodes through optical diffractions. Weights of connection strength of the neurons are determined based on diffractive modulation. Each optoelectronic neuron is configured to perform an optical field summation of weighted inputs and generate a unit output by applying a complex activation to an optical field occurring naturally in a photoelectronic conversion. Each neuron is a programmable device.

NetCast is an optical neural network architecture that circumvents constraints on deep neural network (DNN) inference at the edge. Many DNNs have weight matrices that are too large to run on edge processors, leading to limitations on DNN inference at the edge or bandwidth bottlenecks between the edge and server that hosts the DNN. With NetCast, a weight server stores the DNN weight matrix in local memory, modulates the weights onto different spectral channels of an optical carrier, and distributes the weights to one or more clients via optical links. Each client stores the activations, or layer inputs, for the DNN and computes the matrix-vector product of those activations with the weights from the weight server in the optical domain. This multiplication can be performed coherently by interfering the spectrally multiplexed weights with spectrally multiplexed activations or incoherently by modulating the weight signal from the weight server with the activations.

A ledger stores chain of custody information for files throughout an enterprise network. By identifying files with a homologous identifier such as a fuzzy hash that permits piecewise evaluation of similarity, the ledger can be used to track a chain of custody over a sequence of changes in content, ownership, and file properties. The ledger can be used, e.g., to evaluate trustworthiness of a file the first time it is encountered by an endpoint, or to apply enterprise policies based on trust.

An apparatus comprises a first mirror; a second mirror; a modulation layer positioned between the first mirror and the second mirror and comprising a plurality of modulation regions; a diffraction layer positioned between the modulation layer and the second mirror, and an input port admitting a light beam into the apparatus. The light beam passes through the diffraction layer and is modulated by the modulation layer to create a first modulated beam before being reflected by the first mirror, the first mirror reflecting the first modulated beam toward the second mirror, the second mirror reflecting the first modulated beam toward the modulation layer to be modulated for at least a second time.

An accelerator for modern convolutional neural networks applies the Winograd filtering algorithm in a wavelength division multiplexing integrated photonics circuit modulated by a memristor-based analog memory unit.