Abstract: Recurrent neural networks, and methods therefor, are provided with diagonal and programming fluctuation to find energy global minima. The method may include storing the matrix of weights in memory cells of a crossbar array of a recursive neural network prior to operation of the recursive neural network; altering the weights according to a probability distribution; setting the weights to non-zero values in at least one of the memory cells in a diagonal of the memory cells in the crossbar array; and operating the recursive neural network.

Abstract: A hardware accelerator including a crossbar array programmed to calculate node values of a neural network, the crossbar array comprising a plurality of row lines, a plurality of column lines, and a memory cell coupled between each combination of one row line and one column line. Also, an energy storing element disposed in the crossbar array between each combination of one row line and one column line and a filter that receives information from the energy storing element and provides new information for each node of the neural network.

Abstract: In one example in accordance with the present disclosure a device is described. The device includes a cross-bar array of memristive elements. Each memristive element has a conductance value. The device also includes a column of offset elements. An offset element is coupled to a row of memristive elements and has a conductance value. The device also includes a number of accumulation elements. An accumulation element is coupled to a column of memristive elements. The accumulation element collects an intermediate output from the column and subtracts from the intermediate output an output from the column of offset elements.

Abstract: Example implementations of the present disclosure relate to in situ transposition of the data values in a memory array. An example system may include a non-volatile memory (NVM) array, including a plurality of NVM elements, usable in performance of computations. The example system may include an input engine to input a plurality of data values for storage by a corresponding plurality of original NVM elements. The example system may further include a transposition engine to direct performance of the in situ transposition such that the plurality of data values remains stored by the original NVM elements.

Abstract: A method for determining a solution to a constrained optimization problem includes programming a weights matrix of a Hopfield network with a first encoded matrix representation of an initial constrained optimization problem. The method also includes employing the Hopfield network to determine a solution to the initial constrained optimization problem. Additionally, the method includes encoding a plurality of constrained optimization problems associated with a target constrained optimization problem into a plurality of encoded matrix representations each of which are a combination of the first and the second encoded matrix representations. The plurality of encoded matrix representations increases in convergence to the second encoded matrix representation of the target constrained optimization problem sequentially. The method further includes re-programming the weights matrix of the Hopfield network in an iterative manner with the plurality of encoded matrix representations.

Abstract: Systems are provided for implementing a hardware accelerator. The hardware accelerator emulate a stochastic neural network, and includes a first memristor crossbar array, and a second memristor crossbar array. The first memristor crossbar array can be programmed to calculate node values of the neural network. The nodes values can be calculated in accordance with rules to reduce an energy function associated with the neural network. The second memristor crossbar array is coupled to the first memristor crossbar array and programmed to introduce noise signals into the neural network. The noise signals can be introduced such that the energy function associated with the neural network converges towards a global minimum and modifies the calculated node values.

Abstract: The nitrate-nitrogen concentration in soil is estimated based on the nitrate-nitrogen 200 nm absorption peak. In one embodiment, a device measures the attenuation spectrum of a soil-extractant mixture over a wavelength range that includes wavelengths in the vicinity of the 200 nm absorption peak (the spectral operating range) and then determines the nitrate-nitrogen concentration based on the attenuation spectrum.

Abstract: A double bias dot-product engine for vector processing is described. The dot product engine includes a crossbar array having N×M memory elements to store information corresponding to values contained in an N×M matrix, each memory element being a memristive storage device. First and second vector input registers including N voltage inputs, each voltage input corresponding to a value contained in a vector having N×1 values. The vector input registers are connected to the crossbar array to supply voltage inputs to each of N row electrodes at two locations along the electrode. A vector output register is also included to receive voltage outputs from each of M column electrodes.

Abstract: Systems and methods are provided for processing an optical signal. An example system may include a source disposed on a substrate and capable of emitting the optical signal. A first waveguide is formed in the substrate to receive the optical signal. A first coupler is disposed on the substrate to receive a reflected portion of the optical signal. A second waveguide is formed in the substrate to receive the reflected portion from the first coupler. A second coupler is formed in the substrate to mix the optical signal and the reflected portion to form a mixed signal. Photodetectors are formed in the substrate to convert the mixed signal to an electrical signal. A processor is electrically coupled to the substrate and programmed to convert the electrical signal from a time domain to a frequency domain to determine a phase difference between the optical signal and the reflected portion.

Abstract: An example finite state machine may include a content-addressable memory. The content-addressable memory may include blocks that respectively store input-terms of the finite state machine. The finite state machine may be configured to, for each received input: select a subset of the blocks of the content addressable memory to enable for searching, the subset being selected based on a current state of the finite state machine, and determine a next state of the finite state machine by searching the currently enabled subset of blocks of the content addressable memory based on the input.

Abstract: A technique includes providing a first set of values to a memristor crossbar array and using the memristor crossbar array to perform a Fourier transformation. Using the memristor crossbar array to perform the Fourier transform includes using the array to apply a Discrete Fourier Transform (DFT) to the first set of values to provide a second set of values.

Abstract: A method of computing in memory, the method including inputting a packet including data into a computing memory unit having a control unit, loading the data into at least one computing in memory micro-unit, processing the data in the computing in memory micro-unit, and outputting the processed data. Also, a computing in memory system including a computing in memory unit having a control unit, wherein the computing in memory unit is configured to receive a packet having data and a computing in memory micro-unit disposed in the computing in memory unit, the computing in memory micro-unit having at least one of a memory matrix and a logic elements matrix.

Abstract: A computer system includes multiple memory array components that include respective analog memory arrays which are sequenced to implement a multi-layer process. An error array data structure is obtained for at least a first memory array component, and from which a determination is made as to whether individual nodes (or cells) of the error array data structure are significant. A determination can be made as to any remedial operations that can be performed to mitigate errors of significance.

Abstract: A method of obtaining a dot product includes applying a programming signal to a number of capacitive memory devices coupled at a number of junctions formed between a number of row lines and a number of column lines. The programming signal defines a number of values within a matrix. The method further includes applying a vector signal. The vector signal defines a number of vector values to be applied to the capacitive memory devices.

Abstract: Disclosed techniques provide for dynamically changing precision of a multi-stage compute process. For example, changing neural network (NN) parameters on a per-layer basis depending on properties of incoming data streams and per-layer performance of an NN among other considerations. NNs include multiple layers that may each be calculated with a different degree of accuracy and therefore, compute resource overhead (e.g., memory, processor resources, etc.). NNs are usually trained with 32-bit or 16-bit floating-point numbers. Once trained, an NN may be deployed in production. One approach to reduce compute overhead is to reduce parameter precision of NNs to 16 or 8 for deployment. The conversion to an acceptable lower precision is usually determined manually before deployment and precision levels are fixed while deployed.

Abstract: A circuit for a neuron of a multi-stage compute process is disclosed. The circuit comprises a weighted charge packet (WCP) generator. The circuit may also include a voltage divider controlled by a programmable resistance component (e.g., a memristor). The WCP generator may also include a current mirror controlled via the voltage divider and arrival of an input spike signal to the neuron. WCPs may be created to represent the multiply function of a multiply accumulate processor. The WCPs may be supplied to a capacitor to accumulate and represent the accumulate function. The value of the WCP may be controlled by the length of the spike in signal times the current supplied through the current mirror. Spikes may be asynchronous. Memristive components may be electrically isolated from input spike signals so their programmed conductance is not affected. Positive and negative spikes and WCPs for accumulation may be supported.

Abstract: A memristive dot-product system for vector processing is described. The memristive dot-product system includes a crossbar array having a number of memory elements. Each memory element includes a memristor. Each memory element includes a transistor. The system also includes a vector input register. The system also includes a vector output register.

Abstract: Examples herein relate to linear transformation accelerators. An example linear transformation accelerator may include a crossbar array programmed to calculate a linear transformation. The crossbar array has a plurality of words lines, a plurality of bit lines, and a memory cell coupled between each unique combination of one word line and one bit line, where the memory cells are programmed according to a linear transformation matrix. The plurality of word lines are to receive an input vector, and the plurality of bit lines are to output an output vector representing a linear transformation of the input vector.

Abstract: Systems and methods are provided for processing an optical signal. An example system may include a source disposed on a substrate and capable of emitting the optical signal. A first waveguide is formed in the substrate to receive the optical signal. A first coupler is disposed on the substrate to receive a reflected portion of the optical signal. A second waveguide is formed in the substrate to receive the reflected portion from the first coupler. A second coupler is formed in the substrate to mix the optical signal and the reflected portion to form a mixed signal. Photodetectors are formed in the substrate to convert the mixed signal to an electrical signal. A processor is electrically coupled to the substrate and programmed to convert the electrical signal from a time domain to a frequency domain to determine a phase difference between the optical signal and the reflected portion.

Abstract: Methods for solving systems of linear equations via utilization of a vector matrix multiplication accelerator are provided. In one aspect, a method includes receiving, from a controller and by the vector matrix multiplication accelerator, an augmented coefficient matrix. The method also comprises implementing Gaussian Elimination using the vector matrix multiplication accelerator by: monitoring, by a register in at least one swap operation, a row order of the augmented coefficient matrix when a first row is swapped with a second row of the augmented coefficient matrix, delivering, by the controller in at least one multiply operation, an analog voltage to a desired row of the augmented coefficient matrix to produce a multiplication result vector, and adding, in at least one add operation, the first row to another desired row of the augmented coefficient matrix to produce an add result vector. Systems and circuits are also provided.