Micron Technology Incorporated

Bryan David Kerstetter

Donald M Morgan

Alan John Wilson

John David Porter

Jeffrey P Wright

Kwang-Ho Cho

Methods, systems, and devices for frequency regulation for memory management commands are described. A memory device may maintain a respective first counter and second counter for each monitoring area of the memory device, where the counters may be incremented for each activate command received for the corresponding monitoring area. If the first counter satisfies a first threshold, an activate command issued to the monitoring area may be ignored. If the second counter fails to satisfy a second threshold, a memory management command issued to the monitoring area may be ignored and the memory device may maintain a value of the second counter, while decrementing the first counter. Alternatively, if the second counter satisfies the second threshold, the memory device may perform a memory management operation associated with a received memory management command and may decrement the first counter and the second counter.

Frequency regulation for memory management commands

A system for providing maximum row active time enforcement for memory devices is disclosed. A host device issues an activate command to activate a memory bank of a plurality of memory banks of a memory. The memory device activates the memory bank and determines whether a precharge command to close the first memory bank has been issued by the host device within a maximum threshold amount of time since issuance of the activate command. If the system determines that the precharge command has been issued by the host device within the threshold, the memory device closes the memory bank via the host-issued precharge command. If, however, the system determines that the precharge command has not been issued by the host device within the threshold, the memory device internally issues a precharge command to close the memory bank to reduce potential data loss and other harmful effects to the memory device.

Maximum row active time enforcement for memory devices

Systems, methods, and apparatuses are provided for wear leveling repair in a memory device. A host is configured to issue a wear leveling command and a repair request to a memory device configured to check source data in a memory of the memory device for errors in response to receiving the wear leveling command from the host, transfer source data in the memory of the memory device to a target page, and repair a source page if the source data includes an error. The memory device is further configured to set a new repair match if a wear leveling repair element was not consumed after receiving the repair request and flush a previous repair match before setting the new repair match if the wear leveling repair element was consumed and a physical address of an incoming repair request is associated with the wear leveling repair element.

Wear leveling repair in a memory device

Systems, methods and apparatuses to log memory errors in memory devices that can perform wear leveling based on physical addresses used in the memory devices to address select memory cells. For example, a controller of a memory sub-system communicates with a memory device installed in the memory sub-system to access memory cells in the memory device. During the communication to access memory cells in the memory device, the controller can determine a memory error at a first address. If the controller transmits the first address to the memory device for memory access at the time of the memory error, the memory device converts the first address to a second address to perform the memory access. The controller can be configured to determine the second address and record, in an error log, the memory error in association with the second address.

Error logging for a memory device with on-die wear leveling

Methods, systems, and devices for memory row-hammer mitigation are described. A memory device may operate based on a scheme that is continuous across power cycles. For example, the memory device may access a region if a value of a counter does not satisfy a threshold value and may access the region if a value of the counter satisfies the threshold value. Upon transitioning power states, the value of the counter may be stored to a non-volatile memory such that it may be accessed when transitioning back to the original power state (e.g., an 'ON' state). Accordingly, the value of the counter may be maintained across power cycles.

Memory row-hammer mitigation

Methods, systems, and devices for techniques for non-volatile data protection are described. As part of a power on operation, a non-volatile memory system may be configured to selectively stored data. For example, the memory system may determine whether a host system is authorized to access data stored in the memory system prior to a power off operation. If the memory system determines that the host system is authorized, the memory device may retain the data. If the memory system determines that the host system is not authorized, the memory system may erase all or a portion of the data. In some cases, the memory system may maintain a retain flag to determine whether the host system is authorized. Additionally or alternatively, the memory system may determine whether a password received from the host system is valid to determine whether the host system is authorized.

Techniques for non-volatile data protection

A method and a device is provided for utilizing unused valid (V) bits residing on a previous command to transmit additional activate information to a memory device. Additional activate information may be transmitted to the memory device without increasing the tRCD time, or increasing the command/address (CA) bus pins, or adding additional circuit area, thereby reducing the impact on the performance of the memory device.

Activate information on preceding command

Embodiments of the disclosure are drawn to apparatuses and methods for automatic soft post-package repair (ASPPR). A memory may receive a row address along with a signal indicating an ASPPR operation, such as a bad page flag being set. A word line engine generates a physical address based on the row address, and ASPPR registers stores the physical address. The time it takes from receiving the row address to storing the physical address may be within the timing of an access operation on the memory such as tRAS. The row address may specify a single page of information. If the bad page flag is set, then a subsequent PPR operation may blow fuses to encode the physical address stored in the ASPPR registers.

Growing community of inventors

Kuna, ID, United States of America

Bryan David Kerstetter