Abstract: Technology is disclosed herein for detecting leaky word lines in a non-volatile storage system. The exact leaky word line may be located very rapidly using a divide and conquer approach. First a determination may be made whether at least one word line in a group such as any of the word lines in a block is leaky. This initial determination can be made very quickly. If no word line in the group is leaky, the search can end. However, responsive to a determination that at least one word line in the group is leaky, a divide and conquer search may be performed in which the group of the word lines is repeatedly divided into smaller sub-groups with selected smaller sub-groups tested for a short circuit until the leaky word line is located.

Abstract: A data storage device includes a memory device and a controller coupled to the memory device. The controller is configured to receive a dataset management (DSM) hint, determine if a second physical memory range associated with a next read operation is located within a threshold number of physical block addresses (PBAs) to a first physical memory range associated with a current read operation, where the next read operation is provided by the DSM hint, and utilize at least a portion of a latency budget associated with the current read operation to optimize a read parameter of the first physical memory range.

Abstract: A system has been described that performs differential temperature compensation based on a differential between the temperature at time of programming and temperature at time of reading for a set of data. Differential temperature compensation is useful for bulk programming/reading (e.g., many pages of data) and/or programming/reading super pages of data (multiple pages residing on different memory die).

Abstract: A data storage device stores files in its memory. The files may be logically fragmented in that various parts of a given file may be located in non-continuous logical addresses, which can be disadvantageous. The host can send a request to the data storage device to reduce such logical fragmentation. For example, the host can send a swap command to the data storage device, in response to which the data storage device swaps the logical addresses of data fragments of two different files. This results in the logical address of one or both of the data fragments being continuous with the logical address of another data fragment of the same file. This logical address swap can take place without physically moving the data in the memory.

Abstract: The present disclosure generally relates to estimating when data to be written will be read or re-written prior to actually writing the data to the memory device. The estimating can be used to smartly route the data to the appropriate memory location at the writing stage or to evict the data from a hot memory location to a colder memory location. To perform the estimating, typical traces or data may be used as may the metadata of the data. Separating data according to the data “temperature” (i.e. the expected access time and frequency), and usage to optimize the SLC partition usage has meaningful impact on several storage metrics such as performance and endurance.

Abstract: Methods and apparatus for management of thermal shutdown in data storage devices are provided. One such data storage device includes a non-volatile memory (NVM) including a thermal shutdown temperature indicative of a maximum temperature at which the NVM will retain data stored therein; and a processor coupled to the NVM, the processor configured to: determine whether a temperature at the NVM exceeds the thermal shutdown temperature; start, responsive to the determination that the temperature at the NVM exceeds the thermal shutdown temperature, a timer; determine, responsive to the timer reaching a preselected duration, whether the temperature at the NVM still exceeds the thermal shutdown temperature; cause, responsive to the determination that the temperature at the NVM still exceeds the thermal shutdown temperature, the NVM to be powered off; and maintain power to components of the data storage device other than the NVM.

Abstract: Certain aspects of the present disclosure provide techniques for performing compute in memory (CIM) computations. A device comprises a CIM module configured to apply analog weights to input data using multiply-accumulate operations to generate an output. The device further comprises a digital weight storage unit configured to store digital weight references, wherein a digital weight reference corresponds to an analog weight of the analog weights. The device also comprises a device controller configured to program the analog weights to the CIM module, cause the CIM module to process the input data, and reprogram one or more analog weights that are degraded. The digital weight references in the digital weight storage unit are populated with values from a host processing device. Degraded analog weights in the CIM module are reprogrammed based on the corresponding digital weight references from the digital weight storage unit without reference to the host processing device.

Abstract: The nonvolatile memory includes a plurality of nonvolatile memory cells configured to store multiple data states; a word line connected to a control gate of at least one of the plurality of non-volatile memory cells; a control gate line to supply a control gate signal; a word line switch connected between the word line and the control gate line to control the potential applied to the word line from the control gate line; and a memory controller circuit. The memory controller circuit is configured to control a word line potential on the word line and a control gate potential on the control gate line and to control a state of the control gate. The memory controller circuit, when the nonvolatile memory transitions to a not-on state, is further configured to turn off the word line switch and to charge the control gate line to a charged potential.

Abstract: An erase process for a group of non-volatile memory cells comprises applying doses of erasing to the group and performing erase verify between pairs of successive doses of erasing. The time needed to complete the erase process can be reduced by optimizing the order of performing erase verify. For example, erase verify can be performed by separately performing erase verify for multiple portions of the group in order from previously determined slowest erasing portion of the group to previously determined fastest erasing portion of the group, and aborting the performing of erase verify prior to completion of erase verify for all of the portions of the group in response to a number erase errors exceeding a limit.

Abstract: Technology is disclosed herein for preventing erase disturb in NAND. Erase voltages are applied to a source line and bit lines associated with selected memory cells, while applying an erase enable voltage to word lines connected to the selected cells. Preventing erase disturb may include raising the channel potential of unselected memory cells to a source line voltage that has a sufficiently low magnitude to not erase the unselected cells given a voltage on word lines connected to the unselected cells. The unselected cells share bit lines with the selected cells and may also share word lines. Preventing erase disturb may also include applying voltages to the select transistors that prevent the erase voltage from passing from the shared bit lines to the channels of the unselected cells. The voltages decrease from the bit lines to the unselected memory cells and may prevent GIDL generation. Current consumption is kept low.

Abstract: Systems and methods are disclosed for providing multilingual media files. In certain embodiments, a data storage device includes a controller configured to: receive a command to write data for a media file to a non-volatile memory, wherein the media file includes one or more frames each including a video frame and a plurality of audio frames associated with a plurality of languages; decode using a decoder a first frame of the media file to determine a logical block address (LBA) for a video frame of the first frame and an LBA for each of a plurality of audio frames of the first frame; write the first frame to the non-volatile memory; and update a logical-to-physical (L2P) table to add information associated with the LBA for the video frame of the first frame and the LBA for each of the plurality of audio frames of the first frame.

Abstract: Anon-volatile memory structure capable of storing weights for layers of a deep neural network (DNN) and perform an inferencing operation within the structure is presented. An in-array multiplication can be performed between multi-bit valued inputs, or activations, for a layer of the DNN and multi-bit valued weights of the layer. Each bit of a weight value is stored in a binary valued memory cell of the memory array and each bit of the input is applied as a binary input to a word line of the array for the multiplication of the input with the weight. To perform a multiply and accumulate operation, the results of the multiplications are accumulated by adders connected to sense amplifiers along the bit lines of the array. The adders can be configured to multiple levels of precision, so that the same structure can accommodate weights and activations of 8-bit, 4-bit, and 2-bit precision.

Abstract: The present disclosure generally relates to methods of operating storage devices. The storage device comprises a controller comprising first random access memory (RAM1), second random access memory (RAM2), and a storage unit divided into a plurality of streams. When a write command is received to write data to a stream, change log data is generated and stored in the RAM1, the previous delta data for the stream is copied from the RAM2 to the RAM1 to be updated with the change log data, and the updated delta data is copied to the RAM2. The delta data stored in the RAM2 is copied to the storage unit periodically. The controller tracks which delta data has been copied to the RAM2 and to the storage unit. During a power failure, the delta data and the change log data are copied from the RAM1 or the RAM2 to the storage unit.

Abstract: Technology is disclosed herein for managing timing parameters when programming memory cells. Timing parameters used sub-clocks in an MLC program mode may also be used for those same sub-clocks in a first SLC program mode. However, in a second SLC program mode a different set of timing parameters may be used for that set of sub-clocks. Using the same set of timing parameters for the MLC program mode and the first SLC program mode saves storage space. However, the timing parameters for the MLC program mode may be slower than desired for SLC programming. A different set of timing parameters may be used for the second SLC program mode to provide for faster program operation. Moreover, the different set of timing parameters used for the faster SLC program mode do not require storage of a separate set of timing parameters.

Abstract: The present disclosure generally relates to methods of operating storage devices. The storage device comprises a controller and a media unit. The capacity of the media unit is divided into a plurality of zones. The controller is configured to make informed use of errors by update zone metadata to indicate one or more first logical block addresses were skipped and to indicate the next valid logical block address is available to store data. The controller is further configured to update zone metadata to recommend to the host device to reset one or more full zones, to recommend to the host device to transition one or more open zones to a full state, to alert the host device that one or more open zones have been transitioned to the full state, and to notify the host device of the writeable zone capacity of each of the plurality of zones.

Abstract: As part of the erase operation for a memory block, one or more post-erase tests can be incorporated into the erase operation to see whether the block has grown any defects. After erasing a block and verifying the erase, the post-erase tests can be performed on the block. As these test involve biasing the block and performing a sensing operation, these post erase tests come with a time penalty. To reduce the associated time penalty and improve memory performance while incorporating the defect tests into the erase process, when biasing the memory array for the post-erase defect tests different ramp rates can be used. In particular, faster ramp rates for bias levels, such those applied to the word lines of the block, are used for the post-erase tests than are used for the same bias level when performing the standard read, program verify, or read verify operations.

Abstract: Technology is disclosed herein for a storage system that reduces the Icc during open block reads. A lower than nominal voltage may be applied to the bit lines during open block reads, which reduces Icc. A nominal bit line voltage may be used during closed block reads. The lower than nominal bit line voltage may be combined with using a lower than nominal read pass voltage (Vread) to unprogrammed word lines during the open block read. The lower than nominal Vread has a lower magnitude than a nominal Vread used during a closed block read. Combining the lower than nominal bit line voltage with the lower than nominal Vread to unprogrammed word lines further reduces Icc during open block reads. The ramp rate of Vread may be relaxed (made slower) during at least some open block reads in combination with the lower than nominal bit line voltage.

Abstract: A method for secure wireless communication executed by at least one processor of a device. A registration certificate is transmitted to the device by a host, the registration certificate including a Long Term Device Key (LTDK) and being generated by a registration server in response to the registration of the host as authorized to connect to the device. In response to receiving a request for securing a Bluetooth connection between the device and the host, the device transmits the LTDK to the host. The device receives, from the host, a connection certificate including connection data for establishing the connection between the host and the device. The connection certificate is signed by a private Long Term Host Key (LTHK) of the host, where the LTHK of the host and the LTDK of the device form a cryptographic Long Term Key pair. The device validates the connection certificate using the LTDK of the device to determine whether the host is authorized to connect to the device.

Abstract: A stage-based frequency optimization for a charge pump achieves a higher area efficiency by operating different stages of the charge pump at their optimized frequency simultaneously, instead of single common frequency, to obtain greater output strength. A first set of stages uses triple well devices as transfer switches and operates at a first, higher frequency. The first stages supply a second set of stages using high voltage devices as transfer switches and operates at a second, lower frequency. The two set of stages are connected through a frequency transition circuit.

Abstract: An apparatus includes a control circuit that is configured to connect to an array of non-volatile memory cells. The control circuit includes a first plurality of data latches configured to connect to non-volatile memory cells of a first plane and a second plurality of data latches configured to connect to non-volatile memory cells of a second plane. The control circuit also includes a shared data transfer data latch configured for transfer of data with the first plurality of data latches and the second plurality of data latches. The shared transfer data latch can be used to transfer data for operations being performed on a first plane to use the data latches on the other plane for storing data for operations on the first plane.