Abstract: A memory device includes a command interface configured to receive write commands from a host device. Additionally, the memory device includes an input buffer configured to buffer a strobe signal from the host device. Furthermore, the memory device includes a first ripple counter and a second ripple counter. The memory device includes command handling circuitry configured to alternatingly start the first ripple counter and the second ripple counter in response to consecutive write commands. The command handling circuitry and/or the first and second ripple counters are configured to suppress a reset of the input buffer if either the first ripple counter or the second ripple counter has not reached a threshold and is still counting.

Abstract: In a compute-in-memory (“CIM”) system, current signals, indicative of the result of a multiply-and-accumulate operation, from a CIM memory circuit are computed by comparing them with reference currents, which are generated by a current digital-to-analog converter (“DAC”) circuit. The memory circuit can include non-volatile memory (“NVM”) elements, which can be multi-level or two-level NVM elements. The characteristic sizes of the memory elements can be binary weighted to correspond to the respective place values in a multi-bit weight and/or a multi-bit input signal. Alternatively, NVM elements of equal size can be used to drive transistors of binary weighted sizes. The current comparison operation can be carried out at higher speeds than voltage computation. In some embodiments, simple clock-gated switches are used to produce even currents in the current summing branches. The clock-gated switches also serve to limit the time the cell currents are on, thereby reducing static power consumption.

Abstract: Systems and methods for non-destructive readback and writeback of an integrated circuit system are provided. Such a system may include an adaptive logic element including a first register pair. The first register pair may include a first register operating at a first frequency and a second register operating at a second frequency. The second frequency may be equal to or lower than the first frequency. The second register may store data from the first register. The adaptive logic element may also include a first clock providing a first clock signal to the first register and a second clock providing a second clock signal. The adaptive logic element may also include a multiplexer that may select the first clock signal or the second clock signal as a clock source for the second register.

Abstract: Systems, methods, and apparatus related to validating data stored in a memory system. In one approach, a DRAM stores data for a host device. A controller that manages the DRAM receives a command from the host device to generate a signature. The controller also receives data from the host device that indicates a region of the DRAM. In response to receiving the command, the controller reads data from the indicated region. A signature is generated by the controller based on the data read from the indicated region. The generated signature is sent to the host device in response to the command.

Abstract: A nonvolatile memory device including a memory cell array, a first voltage generator configured to generate a word line operating voltage for each word line of the memory cell array, a second voltage generator configured to generate a bit line operating voltage of the memory cell array, and a temperature unit configured to determine, from a temperature range table, a temperature range for a temperature code according to a real-time temperature of the memory cell array, and to adjust a power supply voltage of the first or second voltage generator based on a selection signal mapped to the determined temperature range.

Abstract: A processor distributes memory timing parameters and data among different memory modules based upon memory access patterns. The memory access patterns indicate different types, or classes, of data for an executing workload, with each class associated with different memory access characteristics, such as different row buffer hit rate levels, different frequencies of access, different criticalities, and the like. The processor assigns each memory module to a data class and sets the memory timing parameters for each memory module according to the module's assigned data class, thereby tailoring the memory timing parameters for efficient access of the corresponding data.

Abstract: A memory sub-system to track charge loss in memory cells and shifts of voltages optimized to read the memory cells. For example, a memory device can measure signal and noise characteristics of a group of memory cells to calculate an optimized read voltage of the group of memory cells. The memory sub-system having the memory device can determine an amount of charge loss in the group of memory cells, using at least the signal and noise characteristics, the optimized read voltage, and/or the bit error rate of data read using the optimized read voltage. The memory sub-system tracks changes in optimized read voltages of memory cells in the memory device based on the amount of charge loss.

Abstract: Disclosed is a memory device which includes a memory cell array including memory cells, data latches connected with a sensing node and storing data in a first memory cell of the memory cells, a sensing latch connected with the sensing node, a temporary storage node, a switch connected between the sensing latch and the temporary storage node and configured to operate in response to a temporary storage node setup signal, a first precharge circuit configured to selectively precharge a first bit line corresponding to the first memory cell depending on a level of the temporary storage node, and a control logic circuit configured to control a dump operation between the data latches, the sensing latch, and the temporary storage node. The control logic circuit performs the dump operation from the data latches to the sensing latch while the first precharge circuit selectively precharges the first bit line.

Abstract: A method of memory reset includes precharging bit lines of a memory array, asserting a signal at a reset node to remove the precharge voltage, and selecting write drivers associated with the bit lines associated with columns of the memory array that contain memory cells to be reset, with the assertion of the signal at the reset node also resulting in application of desired logic states to inputs of the selected write drivers to cause those selected write drivers to change a logic state of the bit lines associated with those write drivers. The method continues with asserting each word line associated with a row of the memory that contains memory cells to be reset to write desired logic states to all of the memory cells of the columns and rows of the memory to be reset during a single clock cycle, and then deasserting those word lines.

Abstract: A first-in-first-out (FIFO) storage structure within an integrated-circuit component is loaded with qualification values corresponding to respective pairs of edges expected within a timing strobe signal transmitted to the integrated-circuit component. The qualification values are sequentially output from the FIFO storage structure during respective cycles of the timing strobe signal and a gate signal is either asserted or deasserted during the respective cycles of the timing strobe signal according to the qualification values output from the FIFO storage structure.

Abstract: A mask read only memory device is provided. Single-transistor memory cells are arranged in rows and columns. Each word line is associated with a corresponding row. Each bit line is associated with a corresponding column. Each first reference line selectively provides a first potential in a first phase and a second potential in a second phase. Each second reference line selectively provides the second potential in the first read phase and the first potential in the second phase. Each memory cell has a gate coupled to a word line, a drain coupled to a bit line and a source terminal either floating, grounded or coupled to one among a first reference line and a second reference line. One of first to fourth logic values is read during a read operation of the memory cell.

Abstract: A method for computation with recurrent neural networks includes receiving an input drive and a recurrent drive, producing at least one modulatory response; computing at least one output response, each output response including a sum of: (1) the input drive multiplied by a function of at least one of the at least one modulatory response, each input drive including a function of at least one input, and (2) the recurrent drive multiplied by a function of at least one of the at least one modulatory response, each recurrent drive including a function of the at least one output response, each modulatory response including a function of at least one of (i) the at least one input, (ii) the at least one output response, or (iii) at least one first offset, and computing a readout of the at least one output response.

Abstract: Disclosed are ferroelectric devices including devices for performing a multiplication of analog input signals and resonators. In one aspect, a ferroelectric nanoelectromechanical device includes a first structural beam, a first input electrode disposed on a first top portion of the first structural beam, and an output electrode. The apparatus further includes a first ferroelectric film disposed on a second top portion of the first input electrode, and a first resistive layer disposed on a third top portion of the first ferroelectric film, wherein a first electrode is positioned at a first end of the first resistive layer and a second electrode is positioned at a second end of the first resistive layer.

Abstract: The disclosure provides a semiconductor storage device, which can shorten the processing time for error detection and correction. The flash memory of the present disclosure has a NAND chip and an ECC chip. The NAND chip has dedicated input and output terminals which can transmit data with the ECC chip, and the ECC chip has a dedicated input and output terminal which can transmit data with the NAND chip. When reading in the NAND chip, the NAND chip transmits the read data containing the parity data to the ECC chip through the dedicated input and output terminals. The ECC chip detect and correct errors in the read data based on the parity data, and the corrected data is transmitted to the controller through the input and output terminals.

Abstract: Systems and methods relate to arranging atoms into 1D and/or 2D arrays; exciting the atoms into Rydberg states and evolving the array of atoms, for example, using laser manipulation techniques and high-fidelity laser systems described herein; and observing the resulting final state. In addition, refinements can be made, such as providing high fidelity and coherent control of the assembled array of atoms. Exemplary problems can be solved using the systems and methods for arrangement and control of atoms.

Abstract: Noise injection procedures implemented on the die of a non-volatile memory (NVM) array are disclosed. In one example, noise is injected into data by adjusting read voltages to induce bit flips while using feedback to achieve a target amount of information degradation. In another example, random data is iteratively combined with itself to achieve a target percentage of random 1s or 0s, then the random data is combined with data read from the NVM array. In other examples, pixels are randomly zeroed out to emulate dead charge coupled device (CCD) pixels. In still other examples, the timing, voltage, and/or current values used within circuits while transferring data to/from latches or bitlines are adjusted outside their specified margins to induce bit flips to inject noise into the data. The noise-injected data may be used, for example, for dataset augmentation or for the testing of deep neural networks (DNNs).

Abstract: A neural processing unit (NPU) for testing a component during runtime is provided. The NPU may include a plurality of functional components including a first functional component and a second functional component. At least one of the plurality of functional components may be driven for calculation of an artificial neural network. Another one of the plurality of functional components may be selected as a component under test (CUT). A scan test may be performed on the at least one functional component selected as the CUT. A tester for detecting a defect of an NPU is also provided. The tester may include a component tester configured to communicate with at least one functional component of the NPU, select the at least one functional component as a CUT, and perform a scan test for the selected CUT.

Abstract: A semiconductor device includes a memory bank. The memory bank includes a plurality N of memory arrays and a local control circuit. Each memory array includes a plurality of bit cells configured to store bits of information and connected between a plurality of bit lines and a plurality of complement bit lines. The local control circuit is configured to pre-charge the bit lines and the complement bit lines at most N?1 memory array at a time. A method of operating the semiconductor device is also disclosed.

Abstract: A calibration circuit includes a first, second and third pull-up units each connected to a first power supply node, and first and second pull-down units each connected to a second power supply node. A first code generator is configured to generate a first code by comparing a voltage of a pad at which the first pull-up unit is connected to an external resistor with a reference voltage, and a second code generator is configured to generate a second code by comparing a voltage of a first intermediate node with the reference voltage and output the second code to the first and second pull-down units. A third code generator is configured to generate a third code by comparing a voltage of a second intermediate node between the second pull-down unit and the third pull-up unit with the reference voltage.

Abstract: A neural processing unit (NPU) for testing a component during runtime is provided. The NPU may include a plurality of functional components including a first functional component and a second functional component. At least one of the plurality of functional components may be driven for calculation of an artificial neural network. Another one of the plurality of functional components may be selected as a component under test (CUT). A scan test may be performed on the at least one functional component selected as the CUT. A tester for detecting a defect of an NPU is also provided. The tester may include a component tester configured to communicate with at least one functional component of the NPU, select the at least one functional component as a CUT, and perform a scan test for the selected CUT.