Abstract: Systems, methods, and apparatus for a memory device that stores a scrub list in a cache used to reduce data traffic to and from a memory array. In one approach, the cache merges the scrub list with cache data. Data in the scrub list can be identified and distinguished from the cache data by adding a one-bit scrub flag to each data entry in the merged cache. In this merged approach, the cache data shares the same memory as the scrub list. Read data that has an error is saved temporarily in this merged cache until the correct value for the data is written back into the memory array.

Abstract: Methods, systems, and devices for shared decoder architecture for three-dimensional memory arrays are described. A memory device may include pillars coupled to an access line using two transistors positioned between the pillar and the access line. The gates of the two transistors may be coupled with respective gate lines coupled with circuitry configured to bias the gate line as part of an access operation for a memory cell coupled with the pillar. In some cases, the circuitry may be positioned between tiles of the memory device, at an end of one or more tiles of the memory device, between word line combs of a tile of the memory device, or a combination thereof.

Abstract: The present disclosure relates to providing multiple error correction code protection levels in memory. One device includes an array of memory cells and an operating circuit for managing operation of the array. The operating circuit comprises an encoding unit configured to generate a codeword for a first error correction code (ECC) protection level and a second ECC protection level, and decoding unit configured to perform an ECC operation on the codeword at the first ECC protection level and the second ECC protection level. The codeword comprises payload data stored in a plurality of memory cells of the array, and parity data associated with the payload data stored in parity cells of the array.

Abstract: Systems, methods, and apparatus for a memory device. In one approach, a memory device selectively enters a streaming mode when accessing memory cells in a memory array. A controller determines for new read operations whether memory cells will be accessed in a streaming mode or in a random mode. First memory cells addressed using a wordline are read by the controller. The wordline is charged to an initial voltage for reading the first memory cells. When in the streaming mode, instead of discharging the wordline after reading the first memory cells, as is done for a random mode, the controller keeps a minimum bias on the wordline and returns the wordline again to the initial voltage for performing a next read operation to read second memory cells. This saves memory device power.

Abstract: Disclosed are techniques for correcting drift accumulation in memory cells. In some aspects, the techniques described herein relate to a memory device including: a memory array, the memory array including a set of memory cells; and a memory controller configured to read data from the memory array, the memory controller configured to: sense a first distribution of the set of memory cells, detect a missing cell in the first distribution, increase a voltage on the missing cell causing the missing cell to be read as part of the first distribution, detect that a second memory cell in a second distribution was read while sensing the first distribution, and mask the second memory cell and mark the second memory cell as belonging to the second distribution.

Abstract: Methods, systems, and devices for generating a balanced codeword protected by an error correction code are described. A memory device may receive data bits for storage. Based on the data bits, the memory device may generate a codeword that includes the data bits, parity bits, and placeholder bits. The memory device may balance the codeword by inverting one or more packets of the codeword. After balancing the codeword, the memory device may store at least a portion of the codeword in memory so that a later operation or a decoding process reveals the packets that were inverted as part of the balancing process. Accordingly, the memory device may re-invert the appropriate packets to recover the original data bits.

Abstract: Methods, systems, and devices for parallel access in a memory array are described. A set of memory cells of a memory device may be associated with an array of conductive structures, where such structures may be coupled using a set of transistors or other switching components that are activated by a first driver. The set of memory cells may be divided into two or more subsets of memory cells, where each subset may be associated with a respective second driver for driving access currents through memory cells of the subset. Two or more of such second drivers may operate concurrently, which may support distributing current or distributing associated circuit structures across a different footprint of the memory device than other different implementations with a single such second driver.

Abstract: Methods, systems, and devices for balanced codewords for reducing a selected state in memory cells are described. A memory device may divide a sequence of data bits into sets of bits associated with different bit-positions in a coding scheme. The memory device may then balance a first codeword that includes the first set of the data bits in the binary domain to reach a target ratio of logic values for the codeword. Using the first codeword and the other set(s) of data bits, the memory device may balance the remaining two states in the state domain to reach an overall target distribution of the three states. The memory device may then generate one or more codeword(s) for the other set(s) of data bits so that the memory device can write all of the codewords to ternary cells.

Abstract: Apparatuses, methods, and systems for performing sense operations in memory are disclosed. The memory can have a group of memory cells, and circuitry can be configured to perform a sense operation on the group, wherein performing the sense operation includes performing a first sense operation in a first polarity on the group of memory cells to determine a quantity of the memory cells of the group that are in a particular data state, and performing a second sense operation in a second polarity on the group of memory cells to determine a data state of the memory cells of the group. The second polarity is opposite the first polarity, and the second sense operation is a count-based sense operation that uses the determined quantity of memory cells in the particular data state as a counting threshold to determine the data state of the memory cells of the group.

Abstract: Methods, systems, and devices for architecture-based power management for a memory device are described. Aspects include operating a first memory bank within a memory device in a first mode and a second memory bank within the memory device in a second mode. The memory device may receive a power down command for the first memory bank while operating the first memory bank in the first mode and the second memory bank in the second mode and switch the first memory bank from the first mode to a first low power mode while maintaining the second memory bank in the second mode. The first low power mode corresponds to less power consumption by the first memory bank than the first mode. In some cases, switching the first memory bank from the first mode to the first low power mode includes deactivating circuitry dedicated to the first memory bank.

Abstract: Methods, systems, and devices for memory array seasoning are described. Some memory cells may have an undesirably high threshold voltage and thus a seasoning operation may be performed on a target memory cell. To season the target memory cell, a bit line and a word line associated with the cell may be activated. Additionally or alternatively, a word line coupled with a second memory cell (e.g., a helper memory cell) that shares the activated bit line may be activated. Accordingly, current flowing across the target memory cell may be increased, which may reduce its threshold voltage.

Abstract: Methods, systems, and devices for fixed weight codewords for ternary memory cells are described. A memory device may generate a codeword from a set of data bits and invert a portion of the codeword so that the codeword is associated with a target distribution of programmable states. After inverting the portion of the codeword, the memory device store the codeword in a set of ternary cells according to a coding scheme. The memory device may read the codeword from the set of ternary cells and select one or more reference voltages for the set of ternary cells based on the target distribution for the codeword and the states of the ternary cells.

Abstract: Methods, systems, and devices for shared decoder architecture for three-dimensional memory arrays are described. A memory device may include pillars coupled to an access line using two transistors positioned between the pillar and the access line. The gates of the two transistors may be coupled with respective gate lines coupled with circuitry configured to bias the gate line as part of an access operation for a memory cell coupled with the pillar. In some cases, the circuitry may be positioned between tiles of the memory device, at an end of one or more tiles of the memory device, between word line combs of a tile of the memory device, or a combination thereof.

Abstract: Methods, systems, and devices for parallel access in a memory array are described. A set of memory cells of a memory device may be associated with an array of conductive structures, where such structures may be coupled using a set of transistors or other switching components that are activated by a first driver. The set of memory cells may be divided into two or more subsets of memory cells, where each subset may be associated with a respective second driver for driving access currents through memory cells of the subset. Two or more of such second drivers may operate concurrently, which may support distributing current or distributing associated circuit structures across a different footprint of the memory device than other different implementations with a single such second driver.

Abstract: Methods, systems, and devices for architecture-based power management for a memory device are described. Aspects include operating a first memory bank within a memory device in a first mode and a second memory bank within the memory device in a second mode. The memory device may receive a power down command for the first memory bank while operating the first memory bank in the first mode and the second memory bank in the second mode and switch the first memory bank from the first mode to a first low power mode while maintaining the second memory bank in the second mode. The first low power mode corresponds to less power consumption by the first memory bank than the first mode. In some cases, switching the first memory bank from the first mode to the first low power mode includes deactivating circuitry dedicated to the first memory bank.

Abstract: Methods, systems, and devices for shared error correction coding (ECC) circuitry are described. For example, a memory device configured with shared ECC circuitry may be configured to receive data at the shared circuitry from either a host device or a set of memory cells of the memory device. The shared circuitry may be configured to generate a set of multiple syndromes associated with a cyclic error correction code, based on the received data. As part of an encoding process, an encoder circuit may generate a set of parity bits based on the generated syndromes. As part of a decoding process, a decoder circuit may generate an error vector for decoding the received data, based on the generated syndromes. The decoder circuit may also correct one or more errors in the received data based on generating the error vector.

Abstract: Methods, systems, and devices related to balancing data are described. Data may be communicated using an original set of bits that may be partitioned into segments. Each of the original set of bits may have a first value or a second value, where a weight of the original set of bits may be based on a quantity of the set of bits that have the first value. If the weight of the original set of bits is outside of a target weight range, a different, encoded set of bits may be used to represent the data, the encoded set of bits having a weight within the target weight range. The encoded set of bits may be identified based an inversion of the original set of bits in a one-at-a-time and cumulative fashion. The encoded set of bits may be stored in place of the original set of bits.

Abstract: Methods, systems, and devices for generating a balanced codeword protected by an error correction code are described. A memory device may receive data bits for storage. Based on the data bits, the memory device may generate a codeword that includes the data bits, parity bits, and placeholder bits. The memory device may balance the codeword by inverting one or more packets of the codeword. After balancing the codeword, the memory device may store at least a portion of the codeword in memory so that a later operation or a decoding process reveals the packets that were inverted as part of the balancing process. Accordingly, the memory device may re-invert the appropriate packets to recover the original data bits.

Abstract: A method to perform data scrub operations by operating a memory device, the memory device comprising a main memory, an internal Error Correction Code (ECC) engine and a scrub memory. The method comprising: receiving a read command; accessing, based on the receiving the read command, a location in the main memory to read data at the location; error correcting data read during the accessing; and storing at the scrub memory information of the location based at least in part on the correcting the data meeting or exceeding an ECC threshold.

Abstract: Methods, systems, and devices for generating a balanced codeword protected by an error correction code are described. A memory device may receive data bits for storage. Based on the data bits, the memory device may generate a codeword that includes the data bits, parity bits, and placeholder bits. The memory device may balance the codeword by inverting one or more packets of the codeword. After balancing the codeword, the memory device may store at least a portion of the codeword in memory so that a later operation or a decoding process reveals the packets that were inverted as part of the balancing process. Accordingly, the memory device may re-invert the appropriate packets to recover the original data bits.