Abstract: A system includes a processor and a memory storing software code and a machine learning (ML) model. The software code is executed to receive contextual data samples each including raw data and a descriptive label, for each contextual data sample: search a database for a data pattern matching the raw data, determine, when the data pattern is detected, whether the data pattern is correlated with an anomalous event, and generate, when the correlation is determined, training data including a label identifying the anomalous event, and the raw data, the data pattern, or both, to provide one of multiple training data samples, wherein the training data samples describe anomalous events corresponding respectively to the raw data, the data pattern, or both. The software code is further executed to train the ML model, using the training data samples, to provide a trained predictive ML model configured to predict the anomalous events.

Abstract: Methods, systems, and devices for techniques for enhanced system performance after retention loss are described. A memory system may program a page of memory cells in response to receiving a power down notification. As part of the programming, the memory system may record an indication of a voltage threshold of the page and power down for a duration of time, during which the memory system may experience retention loss. Upon powering on, the memory device may compare the voltage threshold of the page to the indication stored prior to powering down and determine a voltage offset for one or more blocks of the memory system. In some cases, the memory system may use the voltage offset to determine a starting bin, and may initiate a bin scan to determine a final bin for the one or more blocks.

Abstract: Described are systems and methods related to a memory block erase protocol. An example system includes a memory device having a memory array including a plurality of memory cells. The system further includes a processing device coupled to the memory device. The processing device is to determine a value of a metric associated with the memory array. Responsive to determine that the value of the metric is below a predetermined threshold, the processing device is further to initiate an erase protocol of the memory device. The processing device is further to erase sets of memory cells associated with one or more memory blocks of the memory array. The processing device is further to receive a programming command directed to the first set of memory cells. The processing device is further to perform a programming operation with respect to a set of memory cells responsive to receiving the programming command.

Abstract: A system can include a memory device with multiple management units, each management unit made up of multiple blocks, and a processing device, operatively coupled with the memory device, to perform various operations including identifying, among the management units, some complete management units and some incomplete management units, as well as performing one type of operation using one or more complete management units. The operations can also include performing another type of operation using one or more incomplete management units where this other type of operation include writing, to one or more incomplete management units, metadata associated with the data stored in complete management units.

Abstract: A processing device in a memory sub-system determines whether a media endurance metric associated with a memory block of a memory device satisfies one or more conditions. In response to the one or more conditions being satisfied, a temperature of the memory block is compared to a threshold temperature range. In response to determining the temperature of the memory block is within the threshold temperature range, the processing device causes execution of a wordline leakage test of a wordline group of a set of wordline groups of the memory block. A result of the wordline leakage test of the target wordline group is determined and an action is executed based on the result of the wordline leakage test.

Abstract: A method includes receiving, by a processing device, signaling indicative of a power cycle (PC) to a memory device (MD) having a first signal indicative of a Power On Operation and a second signal indicative of a Power Off Operation, and determining an Average Power On Time (APOT) of the MD based, at least in part, on a quantity of power cycles (n) to the MD over a predetermined time interval (PTI), and for each PC over the PTI, an amount of time between receipt of the first signal and the second signal. A sum of each of the amount of time between receipt of the first signal and the second signal in the PTI provides a total power on time (T) to the MD, and the APOT is equal to T/n. When the APOT is less than (<) a threshold APOT value, determining a frequency at which to perform media scan operations and performing media scan operations involving the MD at the determined frequency.

Abstract: A replacement block pool for a memory device is established. The replacement block pool comprises one or more valid blocks from a set of valid blocks in the memory device determined based on a constraint defining a minimum number of valid blocks for the memory device. A grown bad block is detected in the memory device. The grown bad block is replaced with a replacement block from the replacement block pool in response to detecting the grown bad block.

Abstract: A method includes receiving, by a memory sub-system, host data to be written to a plurality of blocks of a memory device associated with a memory sub-system, where each of the plurality of blocks are coupled to one of a plurality of word lines of the memory device. The method can further include generating parity data for each word line of the block; dividing the parity data into one of either a first word line parity set or a second word line parity set; generating a reduced parity data set with exclusive or parity values for the first word line parity set and for the second word line parity set; and writing the reduced parity data set in the memory sub-system.

Abstract: A method includes performing a first read operation involving a set of memory cells using a first voltage, determining a quantity of bits associated with the set of memory cells based on the first read operation, performing a second read operation involving the set of memory cells using a second voltage that is greater than the first voltage when the quantity of bits is above a threshold quantity of bits for the set of memory cells, and performing the second read operation involving the set of memory cells using a third voltage that is less than the first voltage when the quantity of bits is below the threshold quantity of bits for the set of memory cells.

Abstract: Methods, systems, and devices for techniques for enhanced system performance after retention loss are described. A memory system may program a page of memory cells in response to receiving a power down notification. As part of the programming, the memory system may record an indication of a voltage threshold of the page and power down for a duration of time, during which the memory system may experience retention loss. Upon powering on, the memory device may compare the voltage threshold of the page to the indication stored prior to powering down and determine a voltage offset for one or more blocks of the memory system. In some cases, the memory system may use the voltage offset to determine a starting bin, and may initiate a bin scan to determine a final bin for the one or more blocks.

Abstract: Methods, systems, and devices for techniques for managing temporarily retired blocks of a memory system are described. In some examples, aspects of a memory system or memory device may be configured to determine an error for a block of memory cells. For example, a controller may determine an existence of the error and may temporarily retire the block. A media management operation may be performed on the temporarily retired block and, depending on one or more characteristics of the error, the temporarily retired block may be enabled or retired.

Abstract: A method includes performing a first read operation involving a set of memory cells using a first voltage, determining a quantity of bits associated with the set of memory cells based on the first read operation, performing a second read operation involving the set of memory cells using a second voltage that is greater than the first voltage when the quantity of bits is above a threshold quantity of bits for the set of memory cells, and performing the second read operation involving the set of memory cells using a third voltage that is less than the first voltage when the quantity of bits is below the threshold quantity of bits for the set of memory cells.

Abstract: Methods, systems, and devices for techniques for retiring blocks of a memory system are described. In some examples, aspects of a memory system or memory device may be configured to determine an error for a block of memory cells. Upon determining the occurrence of the error, the memory system may identify one or more operating conditions associated with the block. For example, the memory system may determine a temperature of the block, a cycle count of the block, a quantity of times the block has experienced an error, a bit error rate of the block, and/or a quantity of available blocks in the associated system. Depending on whether a criteria associated with a respective operating condition is satisfied, the block may be enabled or retired.

Abstract: Methods, systems, and devices for techniques for managing temporarily retired blocks of a memory system are described. In some examples, aspects of a memory system or memory device may be configured to determine an error for a block of memory cells. For example, a controller may determine an existence of the error and may temporarily retire the block. A media management operation may be performed on the temporarily retired block and, depending on one or more characteristics of the error, the temporarily retired block may be enabled or retired.

Abstract: A method includes receiving, by a processing device, signaling indicative of a power cycle (PC) to a memory device (MD) having a first signal indicative of a Power On Operation and a second signal indicative of a Power Off Operation, and determining an Average Power On Time (APOT) of the MD based, at least in part, on a quantity of power cycles (n) to the MD over a predetermined time interval (PTI), and for each PC over the PTI, an amount of time between receipt of the first signal and the second signal. A sum of each of the amount of time between receipt of the first signal and the second signal in the PTI provides a total power on time (T) to the MD, and the APOT is equal to T/n. When the APOT is less than (<) a threshold APOT value, determining a frequency at which to perform media scan operations and performing media scan operations involving the MD at the determined frequency.

Abstract: An apparatus can include a block program erase count (PEC) component. The block PEC component can monitor a quantity of program erase counts (PECs) for each particular type of block of a non-volatile memory array. The block PEC component can further determine which block of the superblock to write host data to next based on the quantity of PECs. The block PEC component can further write host data to the determined block.

Abstract: A method includes writing, to a first sub-set of memory blocks of a first plane associated with a memory device, first data corresponding to recovery of an uncorrectable error and writing, to a first sub-set of memory blocks of a second memory plane associated with the memory device, second data corresponding to recovery of the uncorrectable error. A relative physical location of the first sub-set of memory blocks of the first memory plane and a relative physical location of the first sub-set of memory blocks of the second memory plane are a same relative physical location with respect to the first memory plane and the second memory plane.

Abstract: A variety of applications can include apparatus and/or methods that provide parity data protection to data in a memory system for a limited period of time and not stored as permanent parity data in a non-volatile memory. Parity data can be accumulated in a volatile memory for data programmed via a group of access lies having a specified number of access lines in the group. A read verify can be issued to selected pages after programming finishes at the end of programming via the access lines of the group. With the programming of the data determined to be acceptable at the end of programming via the last of the access lines of the group, the parity data in the volatile memory can be discarded and accumulation can begin for a next group having a specified number of access lines. Additional apparatus, systems, and methods are disclosed.

Abstract: A method is described that includes determining, by a memory subsystem controller of a memory device, a number of memory cells from a set of memory cells that are in a programmed state. The memory subsystem controller further compares the number of memory cells from the set of memory cells that are in the programmed state to a proximity disturb threshold and in response to determining that the number satisfies the proximity disturb threshold, performs a remediation operation on user data stored in the set of memory cells.

Abstract: A method includes receiving, by a processing device, signaling indicative of a power cycle (PC) to a memory device (MD) having a first signal indicative of a Power On Operation and a second signal indicative of a Power Off Operation, and determining an Average Power On Time (APOT) of the MD based, at least in part, on a quantity of power cycles (n) to the MD over a predetermined time interval (PTI), and for each PC over the PTI, an amount of time between receipt of the first signal and the second signal. A sum of each of the amount of time between receipt of the first signal and the second signal in the PTI provides a total power on time (T) to the MD, and the APOT is equal to T/n. When the APOT is less than (<) a threshold APOT value, determining a frequency at which to perform media scan operations and performing media scan operations involving the MD at the determined frequency.