Abstract: A method comprises forming a first insulation layer on an upper surface of a semiconductor substrate, forming a first conductive layer on the first insulation layer, and forming a compound insulation layer on the first conductive layer, wherein the compound insulation layer comprises a nitride sublayer between a lower oxide sublayer and an upper oxide sublayer. A second insulation layer is formed on the compound insulation layer. A trench is formed that extends through the second insulation layer, the compound insulation layer, the first conductive layer, the first insulation layer, and into the semiconductor substrate. The trench is filled with fill insulation material. The second insulation layer and an upper portion of the fill insulation material are removed. A second conductive layer is formed on the compound insulation layer, and on the fill insulation material in the trench.

Abstract: Numerous examples are disclosed of a masking circuit for inputs and outputs in a neural network array. In one example, a system comprises a neural network array comprising a plurality of non-volatile memory cells arranged into rows and columns; and row circuits for respective rows in the neural network array, the row circuits comprising a masking circuit to prevent an application of a sparse input to one or more rows in the array when a condition is satisfied.

Abstract: Numerous examples are disclosed for performing calibration of various electrical parameters in a deep learning artificial neural network. In one example, a system comprises a digital-to-analog converter for receiving an input of k bits and generating a first analog output, a mapping scalar for converting the first analog output into a second analog output, and an analog-to-digital converter for generating an output of n bits from the second analog output, where n is a different value than k.

Abstract: In one example, a method comprises precharging at least one input to or output from a memory array to a supply voltage; performing at least one operation by the memory array requiring the precharged at least one input to or output from the memory array to have a voltage lower than the supply voltage; and discharging the precharged at least one input to or output from the memory array to the voltage lower than the supply voltage in response to the at least one operation.

Abstract: Numerous embodiments of output circuitry for an analog neural memory in a deep learning artificial neural network are disclosed. In some embodiments, a common mode circuit is used with differential cells, W+ and W?, that together store a weight, W. The common mode circuit can utilize current sources, variable resistors, or transistors as part of the structure for introducing a common mode voltage bias.

Abstract: In one example, a system comprises a vector-by-matrix multiplication array comprising non-volatile memory cells organized into rows and columns; a plurality of word lines coupled respectively to rows of the vector-by-matrix multiplication array; and a word line driver coupled to the plurality of word lines, the word line driver comprising a plurality of select transistors coupled to a common control line and the plurality of word lines, and a plurality of bias transistors coupled to the plurality of select transistors and capable of providing a bias voltage to a single select transistor in the plurality of select transistors or to all of plurality of select transistors in response to control signals.

Abstract: Numerous examples of a precision tuning algorithm and apparatus are disclosed for precisely and quickly depositing the correct amount of charge on the floating gate of a non-volatile memory cell within a vector-by-matrix multiplication (VMM) array in an artificial neural network. In one example, a method for performing a read or verify operation in a vector-by-matrix multiplication system comprising an input function circuit, a memory array, and an output circuit block is disclosed, the method comprising receiving, by the input function circuit, digital bit input values; converting the digital input values into an input signal; applying the input signal to control gate terminals of selected cells in the memory array; and generating, by the output circuit block, an output value in response to currents received from the memory array.

Abstract: Numerous examples are disclosed of verification circuitry and associated methods in an artificial neural network. In one example, a system comprises a vector-by-matrix multiplication array comprising a plurality of non-volatile memory cells arranged in rows and columns, the non-volatile memory cells respectively capable of storing one of N possible levels corresponding to one of N possible currents, and a plurality of output blocks to receive current from respective columns of the vector-by-matrix multiplication array and generate voltages during a verify operation of the vector-by-matrix multiplication and generate digital outputs during a read operation of the vector-by-matrix multiplication.

Abstract: Numerous examples are disclosed of an artificial neural network that comprises vector-by-matrix multiplication arrays utilizing analog outputs. In one example, a system comprises a vector by matrix multiplication array comprising a plurality of non-volatile memory cells arranged in rows and columns; and an output circuit to receive a respective neuron current from respective columns of the vector by matrix multiplication array and to generate a respective output voltage, the output circuit comprising a neuron scalar to generate a scaled current from the received respective neuron current, and a current-to-voltage converter to convert the scaled current into the respective output voltage.

Abstract: Examples of programming circuits and methods are provided. In one example, an adjustable programming circuit comprises a first adjustable voltage divider; a second adjustable voltage divider; a first operational amplifier, wherein an output terminal of the first operational amplifier provides a first programming voltage; and a second operational amplifier, wherein the first input terminal of the second operational amplifier is coupled to the output terminal of the second operational amplifier and the first input terminal of the second operational amplifier is coupled to the second output terminal of the first adjustable voltage divider.

Abstract: In one example, a method comprises programming a plurality of analog neural non-volatile memory cells in an array of analog neural non-volatile memory cells to store one of N different values, where Nis a number of different levels that can be stored in any of the analog neural non-volatile memory cells; measuring a current drawn by the plurality of analog neural non-volatile memory cells; comparing the measured current to a target value; and identifying the plurality of the analog neural non-volatile memory cells as bad if the difference between the measured value and the target value exceeds a threshold.

Abstract: Examples for ultra-precise tuning of a selected memory cell are disclosed. In one example, a method of programming a first memory cell in a neural memory to a target value is disclosed, the method comprising programming a second memory cell by applying programming voltages to terminals of the second memory cell; and determining if an output of the first memory cell has reached the target value.

Abstract: Numerous embodiments are disclosed for an output circuit for an analog neural memory in a deep learning artificial neural network. In some embodiments, an output block receives current from a W+ bit line and current from an associated W? bit line, and the output block generates an output signal that is a differential signal in certain embodiments and is a single ended signal in other embodiments.

Abstract: A first example comprises programming a memory cell to store a value; applying a series of currents of increasing size to a bit line of the memory cell; and measuring a voltage of a control gate terminal of the memory cell to determine a bias. A second example comprises programming a memory cell to store a value; applying a predetermined current to a bit line of the memory cell; and measuring a voltage of a control gate terminal of the memory cell to determine a bias.

Abstract: Numerous examples are disclosed for performing calibration of various electrical parameters in a deep learning artificial neural network. In one example, a method comprises adjusting a bias voltage applied to one or more non-volatile memory cells in an artificial neural network, performing a performance target check on the one or more non-volatile memory cells in the artificial neural network, and repeating the adjusting and performing until the performance target check indicates an electrical parameter is within a predetermined range.

Abstract: A memory device includes a SOI substrate comprising bulk silicon, an insulation layer vertically over the bulk silicon, and a silicon layer vertically over the insulation layer. A memory cell includes source and drain regions formed in the bulk silicon with a channel region of the bulk silicon extending therebetween, and a floating gate which includes a first portion of the silicon layer disposed vertically over and insulated from a first portion of the channel region by the insulation layer. The first portion of the silicon layer is epitaxially thickened or a layer of polysilicon is formed on the first portion of the silicon layer. A select gate is disposed vertically over and insulated from a second portion of the channel region. A control gate is disposed vertically over and insulated from the floating gate. An erase gate is disposed vertically over and insulated from the source region.

Abstract: A method of forming memory cells, high voltage devices and logic devices on fins of a semiconductor substrate's upper surface, and the resulting memory device formed thereby. The memory cells are formed on a pair of the fins, where the floating gate is disposed between the pair of fins, the word line gate wraps around the pair of fins, the control gate is disposed over the floating gate, and the erase gate is disposed over the pair of fins and partially over the floating gate. The high voltage devices include HV gates that wrap around respective fins, and the logic devices include logic gates that are metal and wrap around respective fins.

Abstract: Examples of improved charge pumps are disclosed. In one example, a system comprises a first charge path comprising a first stage to boost an input voltage and a second stage to boost a voltage received from the first stage of the first charge path; and a second charge path comprising a first stage to boost an input voltage and a second stage to boost a voltage received from the first stage of the second charge path; wherein an output of the second stage of the first charge path is coupled to the first stage of the second charge path and an output of the second stage of the second charge path is coupled to the first stage of the first charge path.

Abstract: Numerous examples are disclosed of systems and methods to implement redundancy. In one example, a system comprises an array of non-volatile memory cells; a redundant array of non-volatile memory cells; and an input block coupled to respective rows in the array and respective rows in the redundant array and comprising row tag registers and redundant row tag registers.

Abstract: Numerous embodiments of input circuitry for an analog neural memory in a deep learning artificial neural network are disclosed.