Abstract: A method and apparatus for video processing on a data storage device. A chip bound architecture includes a CMOS coupled to one or more NAND die, the CMOS including one or more processors, memories, and error correction code (ECC) engines capable of processing video data. According to certain embodiments, macroblocks are correlated between two I-frames, including motion vectors to define different locations of correlated macroblocks. A P-frame may be determined from a previous I-frame and its correlated macroblocks and motion vectors, while a B-frame may be determined from two or more adjacent I-frames with concomitant macroblocks and motion vectors, as well as P-frames associated with an adjacent I-frame.

Abstract: A method and apparatus for systems and methods for digital signal processing (DSP) in a non-volatile memory (NVM) device comprising CMOS coupled to NVM die, of a data storage device. According to certain embodiments, one or more DSP calculations are provided by a controller to the CMOS components of the NVM, that configure one or more memory die to carry out atomic calculations on the data resident on the die. The results of calculations of each die are provided to an output latch for each die, back-propagating data back to the configured calculation portion as needed, otherwise forwarding the results to the controller. The controller aggregates the results of DSP calculations of each die and presents the results to the host system.

Abstract: A data storage device includes a non-volatile memory (NVM) device and a controller coupled to the NVM device. The controller is configured to create a bad block table that tracks bad blocks of the NVM device, send the bad block table to a host memory location, and check the bad block table to determine whether a block to be read or written to is bad. The controller is further configured to request information on a bad block from the bad block table located in the host memory location, determine that the requested information is not available from the host memory location, and retrieve the requested information from a location separate from the host memory location. A sum of the times to generate a request to check the flat relink table, execute the request, and retrieve the requested information is less than a time to process a host command.

Abstract: A storage system allocates single-level cell (SLC) blocks in its memory to act as a write buffer and/or a read buffer. When the storage system uses the SLC blocks as a read buffer, the storage system reads data from multi-level cell (MLC) blocks in the memory and stores the data in the read buffer prior to receiving a read command from a host for the data. When the storage system uses the SLC blocks as a write buffer, the storage system retains certain data in the write buffer while other data is flushed from the write buffer to MLC blocks in the memory.

Abstract: A storage system allocates single-level cell (SLC) blocks in its memory to act as a write buffer and/or a read buffer. When the storage system uses the SLC blocks as a read buffer, the storage system reads data from multi-level cell (MLC) blocks in the memory and stores the data in the read buffer prior to receiving a read command from a host for the data. When the storage system uses the SLC blocks as a write buffer, the storage system retains certain data in the write buffer while other data is flushed from the write buffer to MLC blocks in the memory.

Abstract: A storage system allocates single-level cell (SLC) blocks in its memory to act as a write buffer and/or a read buffer. When the storage system uses the SLC blocks as a read buffer, the storage system reads data from multi-level cell (MLC) blocks in the memory and stores the data in the read buffer prior to receiving a read command from a host for the data. When the storage system uses the SLC blocks as a write buffer, the storage system retains certain data in the write buffer while other data is flushed from the write buffer to MLC blocks in the memory.

Abstract: The present disclosure generally relates to data storage devices, such as solid state drives (SSDs). A read threshold calibration operation is utilized to generate a calibrated read threshold for one or more voltage states of a cell of a MLC memory. A single-level cell (SLC) read is then executed to sense the ratio of bit values at the read thresholds of the voltage states, where SLC read refers to reading at a single read threshold, rather than to the cell type. The sensing results in a binary page with certain statistics of 1's and 0's. The ratio of 1's (or 0's) in the binary page is used to determine a deviation from the expected ratio, where the deviation is used to adjust the calibrated read threshold to match the voltage states of the MLC memory.

Abstract: A data storage device including a non-volatile memory device including one or more non-volatile memory sets and one or more endurance groups. Each of the endurance groups includes at least one of the non-volatile memory sets. The data storage device includes a controller coupled to the non-volatile memory device. The controller is configured to receive a pending command message from a host interface, where the received pending command message includes a command configured to be executed by a first endurance group of the number of endurance groups. The controller is further configured to determine an assigned command slot for storing the command, where the assigned command slot is selected form one of a private command slot pool associated with the first endurance group or a shared command slot pool, fetch the command from the host device, and store the fetched command in the assigned command slot.

Abstract: Methods and apparatus for managing and optimizing data storage devices that include non-volatile memory (NVM) are described. One such method involves deriving a hint for one or more logical block addresses (LBAs) of a storage device based on information received from a host device and/or physical characteristics of the storage device, such as LBAs that are invalidated together; grouping the LBAs into one or more clusters of LBAs based on the derived hint and a statistical analysis of the physical characteristics of the storage devices; allocating available physical block addresses (PBAs) in the storage device to one of the LBAs based on the one or more clusters of LBAs to achieve optimization of a data storage device.

Abstract: The present disclosure generally relates to efficient data transfer management of zone-append commands for a zoned namespace (ZNS). The ZNS storage device comprises a memory device having a plurality of memory dies, and a controller coupled to the memory device The controller receives a plurality of zone append commands, each zone append command being associated with a zone identification identifying a zone of a plurality of zones, and fetches and aggregates data associated with each zone append command by the zone identification in an append write buffer. The aggregated data is written to the memory device upon the aggregated data for each zone reaching a predetermined programming chunk size, or to a temporary buffer if the predetermined write size is not met. Each zone uses a separate channel when sending the aggregated data for programming to the memory device, allowing multiple channels to be utilized in parallel.

Abstract: An apparatus includes a non-volatile memory and a controller coupled to the non-volatile memory. The controller includes an interface configured to send first data to be stored to the non-volatile memory. The controller further includes a control circuit configured to generate updated control information based on storing of the first data to the non-volatile memory. The interface is further configured to concurrently send second data and the updated control information to be stored at the non-volatile memory.

Abstract: The present disclosure generally relates to improved foggy-fine programming. Rather than initially writing to SLC and then later performing a foggy write to QLC with the data read from SLC and then a fine write to QLC with data re-read from SLC, the foggy write to QLC can be performed in parallel to the initial writing to SLC using the same buffer. Once the foggy write to QLC has completed, and the writing to SLC has also completed, the data buffer can be released. The data written in SLC is then be read from SLC and passes through a relocation buffer for the first and only time to then be written using fine programming to QLC. Thus, the data only passes through the relocation buffer one time and the relocation buffer can be freed to usage after only one pass of the data therethrough.

Abstract: The present disclosure generally relates to efficiently relocating data within a data storage device. By implementing an error correction code (ECC) module in a complementary metal oxide semiconductor (CMOS) chip for each memory die within a memory array of a memory device, the data can be relocated more efficiently. The ECC decodes the codewords at the memory die. The metadata is then extracted from the decoded codewords and transferred to a controller of the data storage device. A flash translation layer (FTL) module at the controller then checks whether the data is valid by comparing the received metadata to FTL tables. If the metadata indicates the data is valid, then the data is relocated.

Abstract: A method and apparatus for enhancing reliability of a data storage device. The storage device controller is configured to convert a typical UBER-type event to an MTBF (FFR) event by converting a data error event into a drive functional failure. In this context, the converted error is not counted as an UBER type event for purposes of determining the reliability of the storage device.

Abstract: The present disclosure generally relates to improved foggy-fine programming. Rather than initially writing to SLC and then later performing a foggy write to QLC with the data read from SLC and then a fine write to QLC with data re-read from SLC, the foggy write to QLC can be performed in parallel to the initial writing to SLC using the same buffer. Once the foggy write to QLC has completed, and the writing to SLC has also completed, the data buffer can be released. The data written in SLC is then be read from SLC and passes through a relocation buffer for the first and only time to then be written using fine programming to QLC. Thus, the data only passes through the relocation buffer one time and the relocation buffer can be freed to usage after only one pass of the data therethrough.

Abstract: For a non-volatile memory system with a multi-plane memory die having a large block size, to be able to more readily accommodate zone-based host data using zones that are of a smaller size that the block size on the memory, the memory system assigns data from different zones to different subsets of the planes of a common memory die. The memory system is configured to accumulate the data from the different zones into different write queues and then assemble the data from the different write zones into pages or partial pages of data that can be simultaneously programmed into memory cells connected to different word lines that are in different sub-blocks of different blocks in the corresponding assigned planes of the die.

Abstract: Exemplary methods and apparatus are provided for implementing a deep learning accelerator (DLA) or other neural network components within the die of a non-volatile memory (NVM) apparatus using, for example, under-the-array circuit components within the die. Some aspects disclosed herein relate to configuring the under-the-array components to implement feedforward DLA operations. Other aspects relate to backpropagation operations. Still other aspects relate to using an NAND-based on-chip copy with update function to facilitate updating synaptic weights of a neural network stored on a die. Other aspects disclosed herein relate to configuring a solid state device (SSD) controller for use with the NVM. In some aspects, the SSD controller includes flash translation layer (FTL) tables configured specifically for use with neural network data stored in the NVM.

Abstract: Apparatus and methods for protecting in-flight data during a fundamental reset of a SSD by a connected host are presented. In embodiments, a controller for the SSD includes an input interface configured to receive commands from the host over a link, and processing circuitry coupled to the input interface. The processing circuitry is configured to, in response to receiving a reset command from the host, reset the link and an address space of the SSD, complete a flush of in-flight data from temporary buffers to non-volatile storage of the SSD, and, during an initialization sequence performed by the host, perform an internal reset. In embodiments, in response to the SSD performing the internal reset, the host's state of the SSD is reset, and the host is caused to re-initialize the link and configure the address space of the SSD.

Abstract: Exemplary methods and apparatus are provided for implementing a deep learning accelerator (DLA) or other neural network components within the die of a non-volatile memory (NVM) apparatus using, for example, under-the-array circuit components within the die. Some aspects disclosed herein relate to configuring the under-the-array components to implement feedforward DLA operations. Other aspects relate to backpropagation operations. Still other aspects relate to using an NAND-based on-chip copy with update function to facilitate updating synaptic weights of a neural network stored on a die. Other aspects disclosed herein relate to configuring a solid state device (SSD) controller for use with the NVM. In some aspects, the SSD controller includes flash translation layer (FTL) tables configured specifically for use with neural network data stored in the NVM.

Abstract: An apparatus includes a plurality of non-volatile memory cells and control circuitry connected to the plurality of non-volatile memory cells. The control circuitry is configured to receive write commands from a host and identify write commands associated with temporary data. In a recovery operation, control data associated with the temporary data is omitted from rebuilt control data.