A memory device includes a memory array and control logic, operatively coupled with the memory array, to perform operations including initiating a read recovery process associated with a block of the memory array. The block includes wordlines at an initial voltage. The operations further include causing an early discharge sequence to be performed on a first set of wordlines of the wordlines during the read recovery process to alleviate latent read disturb. The early discharge sequence includes ramping the first set of wordlines from the initial voltage to a ramping voltage while maintaining a second set of wordlines of the wordlines at the initial voltage.

A method of programing a memory device having a plurality of memory cell groups where each of the memory cell group includes N non-volatile memory cells, where N is an integer greater than or equal to 2. For each memory cell group, the method includes programming each of the non-volatile memory cells in the memory cell group to a particular program state, performing multiple read operations on each of the non-volatile memory cells in the memory cell group, identifying one of the non-volatile memory cells in the memory cell group that exhibits a lowest read variance during the multiple read operations, deeply programming all of the non-volatile memory cells in the memory cell group except the identified non-volatile memory cell, and programming the identified non-volatile memory cell in the memory cell group with user data.

Technology is disclosed for an efficient read NAND memory cells while mitigating read disturb. In an aspect, a read sequence includes a read spike that removes residual electrons from the NAND channels, followed by reading multiple different groups of memory cells, followed by a channel clean operation. The read spike and channel clean mitigate read disturb. The read spike and channel clean each take a significant amount of time to perform. However, since multiple groups of memory cells are read between the read spike and channel clean this time is essentially spread over the reading of multiple groups, thereby improving the average time to read a single group of memory cells. In one aspect, reading the multiple different groups of memory cells includes reading one or more pages from each of the groups of memory cells. In one aspect, each group is in a different sub-block.

Apparatus and methods are described to reduce program disturb for a memory string with a partial select gate drain, which is partially cut by a shallow trench. The memory string with a partial select gate drain is linked with a neighboring full select gate drain that during its programming can casuse a program disturb in the memory string with a partial select gate drain. The bias voltage applied to the selected full select gate drain can be controlled from a high state for low memory program states to a lower state for the high memory program states. The high data states may cause program disturb. The reduction in the bias voltage can match a reduction in the bias voltage applied to the bit lines to reduce the program disturb while providing adequate signal to program the high states on the memory string of the full select gate drain.

Systems, apparatuses and methods may provide for technology that determines a write-to-write delay with respect to a memory cell, wherein one or more neighboring cells are adjacent to the memory cell and controls a write disturb refresh rate of the one or more neighboring cells based on the write-to-write delay. In one example, the technology increments a write counter corresponding to the memory cell by a first value if the write-to-write delay exceeds a delay threshold and increments the write counter by a second value if the write-to-write delay does not exceed the delay threshold, wherein the second value is greater than the first value, and wherein the write disturb refresh rate is controlled based on the write counter.

A method for performing a read of a flash memory includes storing configuration files for a plurality of RRD-compensating RNNs. A current number of PE cycles for a flash memory are identified and TVSO values are identified corresponding to the current number of PE cycles. A current retention time and a current number of read disturbs for the flash memory are identified. The configuration file of the RRD-compensating RNN corresponding to the current number of PE cycles, the current retention time and current number of read disturbs is selected and is loaded into a neural network engine to form an RNN core in the neural network engine. A neural network operation of the RNN core is performed to predict RRD-compensated TVSO values. The input to the neural network operation includes the identified TVSO values. A read of the flash memory is performed using the predicted RRD-compensated TVSO values.

Numerous embodiments are disclosed of a non-volatile memory cell array formed in a p-well, which is formed in a deep n-well, which is formed in a p-substrate. During an erase operation, a negative voltage is applied to the p-well, which reduces the peak positive voltage required to be applied to the cells to cause the cells to erase.

A three-dimensional (3D) nonvolatile memory device includes a cell string. The cell string includes a pillar structure comprising a ground selection transistor, a plurality of memory cells, and a string selection transistor stacked vertically over a substrate. The memory cells comprise a first cell group and a second cell group stacked on the first cell group, and a horizontal width of at least a portion of the pillar structure decreases in a depth direction towards the substrate. A method of programming the memory device includes initializing a channel of a memory cell of the first cell group of the cell string through the ground selection transistor of the pillar structure, and then applying a program voltage to the memory cell of the pillar structure of the cell string.

A semiconductor device includes a transmission and reception circuit and a control circuit. The transmission and reception circuit transmits and receives a signal to and from a semiconductor memory device. The control circuit acquires threshold voltage distribution information of a memory element connected to a word line for read disturb detection to which a second voltage higher than a first voltage applied to an adjacent word line adjacent to a read target word line during a read operation is applied and determines an influence of read disturb based on the threshold voltage distribution information.

In certain aspects, a memory device includes a first memory string including a first drain, a first drain select gate (DSG) transistor, first memory cells, and a first drain dummy transistor between the first drain and the first DSG transistor. The memory device also includes a first bit line coupled to the first drain, a first drain dummy line coupled to the first drain dummy transistor, a first DSG line coupled to the first DSG transistor, word lines respectively coupled to the first memory cells, and a peripheral circuit configured to perform a program operation on a target memory cell of the first memory cells coupled to a selected word line of the word lines. To perform the program operation, the peripheral circuit includes a bit line driver coupled to the first bit line and configured to apply a first bit line voltage to select the first bit line, and a word line driver coupled to the first drain dummy line and the first DSG line and configured to apply a DSG voltage to the first DSG line to turn on the first DSG transistor, and apply a drain dummy line voltage to the first drain dummy line to turn on the first drain dummy transistor. The drain dummy line voltage is greater than the DSG voltage.