Abstract: Apparatuses and methods can be related to implementing adjustable data protection schemes using artificial intelligence. Implementing adjustable data protection schemes can include receiving failure data for the plurality of memory devices and receiving an indication of a failure of a stripe of the plurality of memory devices based on the failure data. Based on failure data, and the indication of the failure of the stripe of the plurality of memory devices, a data protection scheme adjustment can be generated for the memory device. The data protection scheme adjustment can be received from the AI accelerator and can be implemented by a plurality of memory devices.

Abstract: Methods, systems, and devices for techniques for enhanced read performance on blocks of memory cells are described. The method may involve selecting a first block of memory cells from a set of blocks of memory cells of a memory system based on a condition of the first block of memory cells being met and setting one or more programming parameters corresponding to the first block of memory cells such that the one or more programming parameters are within a threshold value of one or more programming parameters corresponding to a second block associated with a storage density different from a storage density of the first block of memory cells. Further, the method may involve performing an operation on the block of memory cells according to the one or more programming parameters.

Abstract: A method includes forming at least a portion of a first superblock using a first subset of blocks from at least one memory die of a memory sub-system and forming at least a portion of a second superblock using a second subset of blocks from the at least one memory die of the memory sub-system.

Abstract: Apparatus and methods are disclosed, including memory devices and systems. Example memory devices, systems and methods include a stack of memory dies, a controller die, and a buffer. Example memory devices, systems and methods include one or more neuromorphic layers logically coupled between one or more dies in the stack of memory dies and a host interface of the controller die.

Abstract: A production host can learn the production state awareness (PSA) modes supported by a memory device and select a particular of one of the supported PSA modes. The memory device can receive host image data from the production host and write the host image data according to the selected PSA mode. The memory device can set a PSA state to load complete after writing the host image data. The memory device can thereby be better situated for being soldered to a memory sub-system.

Abstract: Methods, systems, and devices for logical-to-physical (L2P) mapping compression techniques are described. A memory system may use an L2P mapping to map logical addresses to physical addresses of the memory system. The L2P mapping may be a hierarchical L2P mapping divided into multiple levels or subsets that are used to identify a physical address corresponding to a logical address. The memory system may write data to a set of physical addresses that are consecutively indexed and may set a flag in an entry of a second-level of the L2P mapping (e.g., of a three-level L2P mapping) to indicate that the entry is associated with a starting physical address of the consecutively indexed physical addresses. The memory system may subsequently read the data starting at the starting physical address based on the flag (e.g., bypassing reading an entry of a lowest-level of the L2P mapping to determine the physical address).

Abstract: Methods, systems, and devices for memory operations are described. A host system may obtain data for writing to a memory system. The host system may send, to the memory system, an indication that the data is to be written to the memory system, and the memory system may remove invalid data at the memory system until the memory system has sufficient resources to store the data. Based on the memory system having sufficient resources, the memory system may delay background operations at the memory system until the data has been written to the memory system. The memory system may also create a restore point based on the memory system having sufficient resources and receiving the data. In other examples, the removal of invalid data at the memory system may be delayed until after the data is written to the memory system.

Abstract: Described apparatuses and methods relate to a read data path for a memory system. The memory system may include logic that receives data and associated metadata from a memory. The logic may perform a reliability check on the data using the associated metadata to determine if the data has an error. If the data is determined not to include an error, the data may be transmitted to a requestor. If the data is determined to include an error, however, a data recovery process may be initiated to recover the data. This may reduce a likelihood the memory system returns corrupted data to a requestor. The memory system may process a different read request at least partially in parallel with the data recovery process to increase throughput or reduce latency. In some cases, the data recovery process may involve one or more techniques related to redundant array of disks (RAID) technology.

Abstract: Methods, systems, and devices related to signal development caching in a memory device are described. In one example, a memory device in accordance with the described techniques may include a memory array, a sense amplifier array, and a signal development cache configured to store signals (e.g., cache signals, signal states) associated with logic states (e.g., memory states) that may be stored at the memory array (e.g., according to various read or write operations). The memory device may also include a controller configured to determine whether data associated with an address of the memory array is stored in one or more cache blocks of the signal development cache. As an example, the memory device may determine whether the data is stored in one or more cache blocks of the signal development cache based on mapping information associated with the address of the memory array.

Abstract: An apparatus includes a processing resource configured to receive a feature vector of a data stream. The feature vector includes a set of feature values. The processing resource is further configured to calculate a set of feature labels based at least in part on the set of feature values to generate a label vector, provide the label vector to another processing resource, and receive a plurality of classifications corresponding to each feature label of the label vector from the other processing resource. The plurality of classifications are generated based at least in part on a respective range of feature values of the set of feature values. The processing resource is configured to then combine the plurality of classifications to generate a final classification of the data stream.

Abstract: The disclosed embodiments relate to logging activities of memory devices and adjusting the operation of a controller based on the activities. In one embodiment, a method comprises monitoring, by a memory device, die temperatures and data sizes of commands issued to the memory device; determining, by the memory device, a target size for a buffer based on the die temperatures and data sizes; and adjusting, by the memory device, a current size of the buffer to meet the target size.

Abstract: Systems, methods, and computer readable media for tracking memory deltas at a cache line granularity. The method includes receiving a base address for a physical memory region, receiving a list of empty log memory buffers associated with a delta logging session, and responsive to determining that a cache line associated with the physical memory region may be in a modified state, storing the modified cache line and metadata associated with the modified cache line in an active log memory buffer referenced by the list of empty log memory buffers. The method also includes determining that the active log memory buffer is full and appending a flag to the active log memory buffer, thereby marking the active log memory buffer as a full log memory buffer. The method also includes storing a list of full log memory buffers, wherein the list is visible to a host processor.

Abstract: The disclosure includes apparatuses, methods, and systems for pre-decoder circuitry. An embodiment includes a memory array including a plurality of memory cells, decoder circuitry coupled to the array and comprising a first and second n-type transistor having a first and second gate, respectively, and pre-decoder circuitry to provide a bias condition for the first and second gate to provide a selection signal to one of the cells. The bias condition comprises a positive voltage for the first gate and a negative voltage for the second gate for a positive memory cell configuration, and zero volts for the first gate and the negative voltage for the second gate for a negative memory cell configuration. The pre-decoder circuitry comprises first pre-decoder circuitry to provide the positive voltage for the first gate and the zero volts for the second gate and second pre-decoder circuitry to provide the negative voltage for the second gate.

Abstract: Methods, systems, and devices for a dynamic row hammering threshold for memory are described. A memory device may implement a dynamic threshold, such as a threshold quantity of activate operations or a row hammering threshold, for a set of multiple rows of the memory device. For example, the memory device may determine a quantity of rows which exceed a row hammering threshold during a refresh duration and a total quantity of activate operations performed across the set of rows during the refresh duration, and may alter the dynamic threshold based on the quantity of rows, the quantity of activate operations, or both. By altering the dynamic threshold, the memory device may decrease a likelihood that a relatively large quantity of refresh operations for rows that are close to being hammered occur within a short time span.

Abstract: A memory device includes a plurality of memory dies, each memory die of the plurality of memory dies comprising a memory array and a power management component, operatively coupled with the memory array. The power management component sends a test value to one or more other power management components on one or more other memory dies of the plurality of memory dies and receives one or more other test values from the one or more other power management components. The power management component compares the test value and the one or more other test values to a set of expected values, and responsive to the test value and the one or more other test values matching the set of expected values, determines that signal connections between the power management component and the one or more other power management components are functional.

Abstract: A computing system (100) having a storage system that includes a storage device (130) and a host device (105), where the host device (105) is configured to issue memory access commands to the storage device (130). The computing system (100) further includes a prediction system (190) comprising processing circuitry that is configured to perform operations that cause the prediction system (190) to identify one or more components of the storage system (918) that limit random rad performance of the storage system (918). The operations further cause the prediction system (190) to obtain characterization data that is indicative of the impact of the one or more components on random read performance and generate a model based on the characterization data to predict random read performance of the storage system (918).

Abstract: An apparatus includes a subword driver configured to drive a subword line, wherein the subword driver includes a transistor coupled to the subword line, a word driver control circuit configured to provide a first control signal and a second control signal, and a word driver configured to receive the first and second control signals, and based on the first control signal provide a driving signal including a plurality of reset pulses to the transistor of the subword driver to activate the transistor a corresponding plurality of times to discharge the subword line, and further provide the driving signal including a transition following the plurality of reset pulses to activate the transistor to further discharge the subword line.

Abstract: Methods for manufacturing semiconductor devices having a flexible reinforcement structure, and associated systems and devices, are disclosed herein. In one embodiment, a method of manufacturing a semiconductor device includes electrically coupling at least one semiconductor die to a redistribution structure on a first carrier. The semiconductor die can include a first surface facing the redistribution structure and a second surface spaced apart from the redistribution structure. The method also includes reducing a thickness of the semiconductor die to no more than 10 ?m. The method further includes coupling a flexible reinforcement structure to the second surface of the at least one semiconductor die.

Abstract: Methods, systems, and devices for centralized error correction circuit are described. An apparatus may include a non-volatile memory disposed on a first die and a volatile memory disposed on a second die (different than the first die). The apparatus may also include an interface controller disposed on a third die (different than the first die and the second die). The interface controller may be coupled with the non-volatile memory and the volatile memory and may include an error correction circuit that is configured to operate on one or more codewords received from the volatile memory.

Abstract: Semiconductor devices having redistribution structures, and associated systems and methods, are disclosed herein. In some embodiments, a semiconductor assembly comprises a die stack including a plurality of semiconductor dies, and a routing substrate mounted on the die stack. The routing substrate includes an upper surface having a redistribution structure. The semiconductor assembly also includes a plurality of electrical connectors coupling the redistribution structure to at least some of the semiconductor dies. The semiconductor assembly further includes a controller die mounted on the routing substrate. The controller die includes an active surface that faces the upper surface of the routing substrate and is electrically coupled to the redistribution structure, such that the routing substrate and the semiconductor dies are electrically coupled to the controller die via the redistribution structure.