Abstract: A semiconductor memory system and an operating method thereof include a plurality of memory devices; and a controller coupled with the memory devices and configured to perform a decoding process to collect at least an output vector associated with an input bit vector, wherein the input bit vector corresponds to a plurality of variable nodes; divide the plurality of variable nodes into a plurality of groups; calculate syndrome of the output vector; and update flipping threshold and flipping indicator of each of the plurality of variable nodes.

Abstract: Techniques are described for performing a check node update (CNU) as part of iterative decoding of a low density-parity check (LDPC) code. The CNU uses a min-sum decoding approach that monitors whether two values received in messages from two variable nodes connected to a check nodes are equal and are the minimum value among the values received by the check nodes from other variable nodes connected thereto. Upon detecting such an event, the minimum value is adjusted by reducing it by an adjustment value to generate an adjusted minimum value. This adjusted minimum value approximates the minimum value that a sum-product algorithm (SPA) decoding approach would have generated. The adjusted minimum value is included in a response message sent from the check node to a variable node. The bit corresponding to that variable node is decoded based on this adjusted minimum value.

Abstract: Decoding method includes calculating cyclic redundancy check (CRC) parity bits for data on-the-fly; performing a low-density parity-check (LDPC) decoding for the data; if it is determined that an iteration of is finished, updating the calculated CRC parity bits to generate CRC parity bits; comparing the generated CRC parity bits with CRC bits included in the data; and terminating the LDPC decoding based on the comparison result.

Abstract: A memory system, a controller including a bit-flipping (BF) decoder and a min-sum (MS) decoder that may be included in the memory system and operating methods thereof in which the controller determines a quality metric as a function of initial syndrome weight and information of the BF decoder after a set number of decoding iterations by the BF decoder in a test period. After the test period, the controller applies the quality metric to each codeword to determine whether to send that codeword first to the BF decoder for decoding or directly to the MS decoder for decoding.

Abstract: A system includes a plurality of memory cells. Each memory cell is programmed to a data state corresponding to one of multiple cell programmed voltages. The memory cells are read to determine a programmed data state of each memory cell. Error correction decoding is performed to determine a corrected data state of each memory cell. The corresponding cell levels, or programmed voltages, are determined based on the programmed data state and the corrected data state. A first error count represents a total number of error cells that have a higher cell level for the corrected data state than the programmed data state. A second error count represents a total number of error cells that have a lower cell level for the corrected data state than the programmed data state. The system is configured to perform a memory operation based on the first error count and the second error count.

Abstract: Adaptive read-threshold schemes for a memory system determine read-threshold with the lowest BER/UECC failure-rates while continuing to serve the host-reads with the required QoS. When it is determined that the QoS or other quality metric is not met for a particular read-threshold, which may be an initial, default, read-threshold, the performance of other read-thresholds are estimated. These estimates may then be used to determine an optimal read-threshold. During the iterative process, selection variables, e.g., how many times, and for which read commands, to use each of the non-default read-thresholds in future read-attempts may be determined on-the-fly.

Abstract: A system includes memory cells arranged in blocks and a memory controller. The memory controller receives a read command to read a first block. The first block can be associated with a first read count and a first read threshold. The first read count is incremented when the first block is read, and when the first read count reaches the read threshold, a read reclaim test is performed. The first read count is set to zero after a power off or a read reclaim operation. When the first read count is zero, an adaptive read threshold is selected based on the number of bit errors. Further, in a read reclaim test, the number of bit errors is tested against an adaptive error threshold to determine whether a garbage collection operation is performed.

Abstract: Techniques are described for performing a bit-flipping decoding scheme on a G-LDPC codeword. In an example, a decoding system uses two syndrome tables. The first syndrome table identifies a predefined syndrome for a component codeword that protects a bit of the G-LDPC codeword. This predefined syndrome is identified based on a location of the bit and is used to update a current syndrome of the component codeword. The second syndrome table identifies one or more bit error locations for the component codeword. The bit error locations are identified from the second syndrome table based on the current syndrome of the component codeword, as updated. In an example, the error locations are used to update a reliability of the bit if its location corresponds to one of the error locations. A bit flipping decision is made for the bit based on its reliability.

Abstract: A memory system, a bit-flipping (BF) low-density parity check (LDPC) decoder that may be included in the memory system and operating methods thereof in which such decoder or decoding has a reduced error floor. Such a BF LDPC decoder is configured using a deep learning framework of trained and training neural networks and data separation that exploits the degree distribution information of the constructed LDPC codes.

Abstract: Techniques are described for encoding information data bits using a low-density parity-check matrix optimized for a Low-Density Parity-Check (LDPC) encoder. In an example, the parity-check matrix includes a first matrix and a second matrix. The second matrix is a square matrix, and is also a block diagonal matrix that includes a set of square submatrices located on the diagonal of the block diagonal matrix. An intermediate vector is generated based on the information data bits and the first matrix, and a parity vector of a codeword is generated based on the intermediate vector and the second matrix.

Abstract: A memory device including a memory having a plurality of memory cells for storing data. The memory device includes a controller communicatively coupled to the memory and configured to organize the data as a plurality of stripes. Each individual stripe of the plurality of stripes includes a plurality of data groups, each of the plurality of data groups stored in the memory using a subset of the plurality of memory cells. Stripe lengths (number of data groups) for individual stripes are determined by the controller based on detecting a condition associated with one or more data groups of the plurality of data groups. At least one data group of the plurality of data groups for each of the individual stripes includes parity data for correcting bit errors associated with the subset of the plurality of memory cells for the individual stripe.

Abstract: A semiconductor memory system and an operating method thereof include a plurality of memory devices; and a memory controller including a feature booster and a linear predictor and coupled with the plurality of memory devices, wherein the controller is configured to collect NAND data from at least 1 data point, and model the collected NAND data with a mixture model, wherein the mixture model includes parameters and at least two latent variables modeled with different distribution modeling, the feature booster is configured to predict the parameters, and the linear predictor is configured to predict feature information.

Abstract: Techniques for codeword decoding are described. In an example, a system accesses information about a block of a storage device of the system. The block includes data lines and redundant lines. The data lines are available to store data and include a bad data line that is unreliable for data storage. The redundant lines include a redundant line that replaces the bad data line, and a free redundant line that remains available after replacement of all bad data lines from the data lines. The information includes an identifier of the bad data line and an identifier of the free redundant line. The system accesses a codeword stored in the block. A portion of the codeword is stored in the free redundant line. The system decodes the codeword based on the identifier of the bad data line and the identifier of the free redundant line.

Abstract: Techniques for improving the bit error rate (BER) performance of an error correction system are described. In an example, the error correction system implements generalized low-density parity-check (GLDPC) encoding and decoding. To generate a GLDPC codeword, the error correction system accesses data blocks. Each data block includes one or more bits. The error correction system also generates a first constituent codeword of the GLDPC codeword. The first constituent codeword encodes at least a data block from the data blocks and has a first length and a first error correction capability. The error correction system also generates a second constituent codeword of the GLDPC codeword. The second constituent codeword encodes at least the data block and has a second length and a second error correction capability. The second length is different from the first length. The second error correction capability is different from the first error correction capability.

Abstract: Memory systems may include a memory storage, and an error correcting code (ECC) unit suitable for determining a number of unsatisfied check nodes of a channel output in a decoding iteration of a decoding process, updating a flipping indicator of a variable node, comparing the flipping indicator of the variable node with a flipping threshold associated with the decoding process, flipping a bit of the variable node when the flipping indicator is greater than the flipping threshold, and ending the decoding process when decoding is determined to be successful or a maximal iteration number is reached.

Abstract: Techniques for improving the latency or processing performance of an error correction system are described. In an example, the error correction system implements LDPC decoding and uses an early termination rule to determine whether the LDPC decoding should be terminated prior to reaching a maximum number of iterations. The early termination rule involves various parameters that relate to the syndrome of the decoded LDPC codeword at each iteration. These parameters include the number of the current decoding iteration and the weight of the syndrome at the current iteration. For example, the early termination rule specifies that the LDPC decoding should be terminated prior to the maximum number of iterations either when the weight of the syndrome is zero, or when the current number of iterations reaches an iteration number threshold and the weight of the syndrome equals or exceeds a checksum threshold.

Abstract: Techniques for improving the bit error rate (BER) performance of an error correction system are described. In an example, the error correction system implements low-density parity-check (LDPC) decoding that uses bit flipping. In a decoding iteration, a feature map is generated for a bit of an LDPC codeword. The bit corresponds to a variable node. The feature map is input to a neural network that is trained to determine whether bits should be flipped based on corresponding feature maps. An output of the neural network is accessed. The output indicates that the bit should be flipped based on the feature map. The bit is flipped in the decoding iteration based on the output of the neural network.

Abstract: Techniques for improving the bit error rate (BER) performance of an error correction system are described. In an example, the error correction system implements generalized low-density parity-check (GLDPC) encoding and decoding. To generate a GLDPC codeword, the error correction system accesses data blocks. Each data block includes one or more bits. The error correction system also generates a first constituent codeword of the GLPDC codeword. The first constituent codeword encodes at least a data block from the data blocks and has a first length and a first error correction capability. The error correction system also generates a second constituent codeword of the GLPDC codeword. The second constituent codeword encodes at least the data block and has a second length and a second error correction capability. The second length is different from the first length. The second error correction capability is different from the first error correction capability.

Abstract: An apparatus of a memory system and an operating method thereof includes a plurality of memory devices; and a controller coupled with the plurality of memory devices, configured to determine a range of read reference voltages having a plurality of read reference voltages, the read reference voltages achieving a minimal rBER; calculate an optimal read reference voltage in accordance with at least the range of read reference voltages; achieve a rBER in accordance with at least the optimal read reference voltage; and execute error correction process with at least the optimal read reference voltage.

Abstract: Techniques for processing bits associated with an “N” multiple level cell NAND flash memory, such as a QLC NAND flash memory, are described. In an example, a system generates a symbol based on the bits. The symbol corresponds to at least two bits. The system encodes the symbol in a non-binary codeword and stores the non-binary codeword in the “N” multiple level cell NAND flash memory based on a mapping between symbols and voltage levels of the “N” multiple level cell NAND flash memory. The system initializes a non-binary decoding procedure based on asymmetric crossover probabilities between the voltage levels. The asymmetric crossover probabilities are defined based on the mapping between the symbols and the voltage level. The system decodes the non-binary codeword based on the non-binary decoding procedure.