Abstract: A matrix multiplication and addition (MAC) operating system is described where different cell currents flowing through different cells in a matrix for different PWM time intervals can be integrated and converted into a numeric value for the cumulative charge. The numeric value of the cumulative charge computed by the ADC is equivalent to MAC operations over multiplicity of cells. The operation is inherently error prone due to parasitic coupling effects from the switching of the memory cells in the array. The presented system minimizes the errors due to parasitic coupling through memory array.

Abstract: In one aspect, an analog to digital converter (ADC) for a multiply-accumulator (MAC) system comprising: an ADC control that receives a VREF and generates a plurality of timing signals to an ADC, and wherein the plurality of timing signals comprises an S1 signal, an S3 signal, an S4 signal, an ECO signal, a CLOCK signal, and a COUNTER<N:0> signal; the ADC that comprises: a pre-charge system comprising a sense capacitor that stores an integrated charge IMAC over a time T and develops voltage VMAC, and wherein the S1 signal defines the pre-charge phase of the sense capacitor.

Abstract: In one aspect, a method for NOR flash cell-array programming in a neural circuit includes the step of erasing a cell array. The method includes the step of programming a set of reference cells of a reference cell array to a target reference threshold voltage (Vt_ref). The method includes the step of generating, with the reference cells, a current or voltage, reference signal. The method includes the step of using the reference signal to bias the neural cells during verification of program state of the neural cells to achieve their respective target threshold voltages (Vt_cell). The method includes the step of programming a set of neural cells of a neural cell array to their respective target threshold voltages.

Abstract: Memory having an array of memory cells might include control logic configured to cause the memory to program each memory cell of a plurality of memory cells whose respective data state is higher than or equal to a first particular data state of a plurality of data states while inhibiting programming of each memory cell of the plurality of memory cells whose respective data state is lower than the first particular data state, and program each memory cell of the plurality of memory cells whose respective data state is lower than or equal to a second particular data state of the plurality of data states after programming each memory cell of the plurality of memory cells whose respective data state is higher than or equal to the first particular data state.

Abstract: The method provides for a low power and a temperature independent analog to digital convertor for systems which use non-volatile cells for forming neurons to be used for neural network applications. The method uses a common counter which can be an up-counter or a down-counter depending on implementation, but in which the source and sink currents to a comparator are changed with temperature by the same percentage as the average bit line current for specific weight distributions programmed in the non-volatile cells forming the neurons. The method uses charge accumulation for detecting the average neuron current.

Abstract: In one aspect, a system for analog in-memory compute for a neural network includes an array of neurons. Each neuron of the array of neurons receives a pulse of magnitude xi, and duration t, wherein a product xi*yi provides a current proportional to the input for a time duration t, which is a charge associated with a particular neuron in response to the input being presented to that particular neuron. A reference cell includes a gate-drain connected flash cell configuration and coupled with the array of neurons. The reference cell is programmed to a pre-determined threshold voltage Vt-ref. The reference cell receives a pre-determined current, Iref, wherein, based on the Iref and a pre-determined threshold voltage Vt-ref, a voltage is created at a drain of the reference cell.

Abstract: Memory devices might include a plurality of memory cell pairs each configured to be programmed to store a digit of data; and control circuitry configured to cause the memory device to compare the stored digit of data of each memory cell pair to a received digit of data, determine whether a match condition or a no-match condition is indicated between the stored digit of data of each memory cell pair and the received digit of data, and deem a match condition to be met between the received digit of data and the stored digits of data of the plurality of memory cell pairs in response to a match condition being determined for a majority of memory cell pairs of the plurality of memory cell pairs.

Abstract: In one aspect, a method of a neuron circuit includes the step of providing a plurality of 2N?1 single-level-cell (SLC) flash cells for each synapse (Yi) connected to a bit line forming a neuron. The method includes the step of providing an input vector (Xi) for each synapse Yi wherein each input vector is translated into an equivalent electrical signal ESi (current IDACi, pulse TPULSEi, etc). The method includes the step of providing an input current to each synapse sub-circuit varying from 20*ESi to (2N?1)*ESi. The method includes the step of providing a set of weight vectors or synapse (Yi), wherein each weight vector is translated into an equivalent threshold voltage level or resistance level to be stored in one of many non-volatile memory cells assigned to each synapse (Yi).

Abstract: In one aspect, a method of a neuron circuit includes the step of providing a plurality of 2N?1 single-level-cell (SLC) flash cells for each synapse (Yi) connected to a bit line forming a neuron. The method includes the step of providing an input vector (Xi) for each synapse Yi wherein each input vector is translated into an equivalent electrical signal ESi (current IDACi, pulse TPULSEi, etc). The method includes the step of providing an input current to each synapse sub-circuit varying from 20*ESi to (2N?1)*ESi. The method includes the step of providing a set of weight vectors or synapse (Yi), wherein each weight vector is translated into an equivalent threshold voltage level or resistance level to be stored in one of many non-volatile memory cells assigned to each synapse (Yi).

Abstract: In one aspect, a method of a neuron circuit includes the step of providing a plurality of 2N-1 single-level-cell (SLC) flash cells for each synapse (Yi) connected to a bit line forming a neuron. The method includes the step of providing an input vector (Xi) for each synapse Yi wherein each input vector is translated into an equivalent electrical signal ESi (current IDACi, pulse TPULSEi, etc). The method includes the step of providing an input current to each synapse sub-circuit varying from 20*ESi to (2N-1)*ESi. The method includes the step of providing a set of weight vectors or synapse (Yi), wherein each weight vector is translated into an equivalent threshold voltage level or resistance level to be stored in one of many non-volatile memory cells assigned to each synapse (Yi).

Abstract: In one aspect, A method for computing neural network computation includes the step of, providing plurality of neurons, coupled with a plurality of inputs, through a plurality of synapses. Each neuron output is given by an equation ?(Xi*Yi)+b. Xi*Yi comprises the ith synapse of the neuron. Xi comprises a set of Xi input vectors. Each Xi input vector is translated into an equivalent electrical signal for an ith corresponding synapse of the plurality of neurons, Yi comprises a set of Yi weight vectors, wherein each Yi weight vector comprises a parameter for the ith corresponding synapse of the plurality of neurons. Each synapse is a sub-system and the sub-system comprises a negative vector neural circuit, a positive vector neural circuit, and a set of four non-volatile memory weight cells for computation. The method includes the step of identifying the input vector x as a positive input vector or a negative input vector.

Abstract: Memory devices might include a plurality of memory cell pairs each configured to be programmed to store a digit of data; and control circuitry configured to cause the memory device to compare the stored digit of data of each memory cell pair to a received digit of data, determine whether a match condition or a no-match condition is indicated between the stored digit of data of each memory cell pair and the received digit of data, and deem a match condition to be met between the received digit of data and the stored digits of data of the plurality of memory cell pairs in response to a match condition being determined for a majority of memory cell pairs of the plurality of memory cell pairs.

Abstract: Memory having an array of memory cells might include control logic configured to cause the memory to program each memory cell of a plurality of memory cells whose respective data state is higher than or equal to a first particular data state of a plurality of data states while inhibiting programming of each memory cell of the plurality of memory cells whose respective data state is lower than the first particular data state, and program each memory cell of the plurality of memory cells whose respective data state is lower than or equal to a second particular data state of the plurality of data states after programming each memory cell of the plurality of memory cells whose respective data state is higher than or equal to the first particular data state.

Abstract: Memory having an array of memory cells might include control logic configured to cause the memory to inhibit memory cells of a first subset of memory cells from programming during each programming pulse of a first plurality of programming pulses and enable those memory cells for programming for at least one programming pulse of a second plurality of programming pulses, inhibit memory cells of a second subset of memory cells from programming during each programming pulse of the second plurality of programming pulses and enable those memory cells for programming for at least one programming pulse of the first plurality of programming pulses, and enable memory cells of a third subset of memory cells for programming during at least one programming pulse of the first plurality of programming pulses and during at least one programming pulse of the second plurality of programming pulses.

Abstract: Memory devices might include control circuitry that, when checking for a match of a stored digit of data and a received digit of data, might be configured to cause the memory device to apply a first voltage level to a control gate of a first memory cell of a memory cell pair, apply a second voltage level different than the first voltage level to a control gate of a second memory cell of that memory cell pair, determine whether that memory cell pair is deemed to be activated or deactivated in response to applying the first and second voltage levels, and deem a match between the stored digit of data and a received digit of data in response, in part, to whether that memory cell pair is deemed to be deactivated.

Abstract: In one aspect, a system for analog in-memory compute for a neural network includes an array of neurons. Each neuron of the array of neurons receives a pulse of magnitude xi, and duration t, wherein a product xi*yi provides a current proportional to the input for a time duration t, which is a charge associated with a particular neuron in response to the input being presented to that particular neuron. A reference cell includes a gate-drain connected flash cell configuration and coupled with the array of neurons. The reference cell is programmed to a pre-determined threshold voltage Vt-ref.

Abstract: Memory devices might include control circuitry that, when checking for a match of a stored digit of data and a received digit of data, might be configured to cause the memory device to apply a first voltage level to a control gate of a first memory cell of a memory cell pair, apply a second voltage level different than the first voltage level to a control gate of a second memory cell of that memory cell pair, determine whether that memory cell pair is deemed to be activated or deactivated in response to applying the first and second voltage levels, and deem a match between the stored digit of data and a received digit of data in response, in part, to whether that memory cell pair is deemed to be deactivated.

Abstract: The method provides for a low power and a temperature independent analog to digital convertor for systems which use non-volatile cells for forming neurons to be used for neural network applications. The method uses a common counter which can be an up-counter or a down-counter depending on implementation, but in which the source and sink currents to a comparator are changed with temperature by the same percentage as the average bit line current for specific weight distributions programmed in the non-volatile cells forming the neurons. The method uses charge accumulation for detecting the average neuron current.

Abstract: Methods include receiving a pattern to be searched in a memory having a plurality of sets of memory elements with each set coupled to a separate data line and corresponding to a same set of bit positions of the pattern. Methods further include receiving a pattern of data to be programmed into a memory, programming a first data state into one memory cell of each cell pair of a plurality of cell pairs of a memory array, and programing a second data state into another memory cell of each cell pair of the plurality of cell pairs for each bit position of the pattern. Memory configured to facilitate such methods include a plurality of cell pairs, each cell pair of the plurality of cell pairs programmed to store a same bit of data corresponding to a particular bit position of a pattern to be searched in the memory.

Abstract: In one aspect, a method for NOR flash cell-array programming in a neural circuit includes the step of erasing a cell array. The method includes the step of programming a set of reference cells of a reference cell array to a target reference threshold voltage (Vt_ref). The method includes the step of generating, with the reference cells, a current or voltage, reference signal. The method includes the step of using the reference signal to bias the neural cells during verification of program state of the neural cells to achieve their respective target threshold voltages (Vt_cell). The method includes the step of programming a set of neural cells of a neural cell array to their respective target threshold voltages.