Abstract: Methods, systems, and devices for opportunistic data movement are described. A memory device may include a non-volatile memory and a volatile memory that operates as a cache for the non-volatile memory. The memory device may receive a write command from a host device. The write command may be associated with a row of a bank in a volatile memory. The memory device may write data associated with the write command to a buffer that is associated with the bank and that is coupled with the volatile memory. And the memory device may communicate the data from the buffer to the volatile memory based on the write command and before a pre-charge command for the row of the bank is received from the host device.

Abstract: Methods, systems, and devices for cache management in a memory subsystem are described. An interface controller may include a first buffer and a second buffer. The interface controller may use the first and second buffers to facilitate operating a volatile memory as a cache for a non-volatile memory. During an access operation, the interface controller may use the buffer to transfer data between the volatile memory, non-volatile memory, and another device. In response to the access operation, the interface controller may use the second buffer to transfer second data from the volatile memory to the non-volatile memory.

Abstract: A data storage device whose controller is configured to apply a hash function to a logical address specified in a received host request to obtain a first portion of the corresponding physical address (e.g., the channel number or channel and die numbers). This feature of the controller enables the L2P table stored in the DRAM associated with the controller to have physical-address entries that contain therein only complementary second portions of the physical addresses, but not the first portions. Such shorter physical-address entries in the L2P table enable a corresponding beneficial reduction in the size of the DRAM and can further be leveraged to have optimized and aligned access to the L2P table in the DRAM.

Abstract: A multi-memory apparatus that uses machine learning is described. The apparatus may include an interface controller, a non-volatile memory, and a volatile memory. The interface controller may cause the apparatus to receive a first command from a host device. The interface controller may cause the apparatus to communicate the first command to a machine learning engine and to circuitry configured to store and manage commands for the non-volatile memory and the volatile memory. The interface controller may further cause the apparatus to communicate a second command generated by the machine learning engine to the circuitry. The second command may be based on information determined by the machine learning engine during a training mode.

Abstract: Methods, systems, and devices for strategic power mode transition in a multi-memory device are described. A controller may receive, from a host device, a command indicating that the controller is to transition a volatile memory and a non-volatile memory from respective deep sleep modes. In a first example, the controller may respond to the command by transitioning the volatile memory to a standby power mode for the volatile memory and transitioning the non-volatile memory to an intermediate power mode for the non-volatile memory that consumes less power than a standby mode for the non-volatile memory. In a second example, the controller may respond to the command by transitioning the volatile memory to the standby power mode for the volatile memory and maintain the non-volatile memory in the deep sleep mode until a condition, such as a miss, occurs.

Abstract: Methods, systems, and devices for power mode control in a multi-memory device are described. An apparatus may include a non-volatile memory and a volatile memory. The apparatus may operate the volatile memory in a first power mode and the non-volatile memory in a second power mode. The apparatus may transition the volatile memory from the first power mode to a third power mode based on a power mode command from a host device. The apparatus may transition the non-volatile memory from the second power mode to a fourth power mode that consumes less power than the second power mode irrespective of the command from the host device and based on a quantity of queued commands for the non-volatile memory being less than a threshold quantity.

Abstract: Methods, systems, and devices for scratchpad memory in a cache are described. A device may operate a portion of a volatile memory in a cache mode having non-deterministic latency for satisfying requests from a host device. The device may monitor a register with an output pin that is associated with the portion and indicative of an operating mode of the portion. Based on or in response to monitoring the output pin, the device may determine whether to change the operating mode of the portion from the cache mode to a scratchpad mode having deterministic latency for satisfying requests from the host device.

Abstract: Methods, systems, and devices for hazard detection in a multi-memory device are described. A device may receive a first command that indicates a first bank address, a first row address, and a first column address. Based on the first bank address, the device may select a buffer for a hazard detection procedure that detects hazardous commands. The device may compare, as part of the hazard detection procedure, the first row address and the first column address from the first command with a second row address and a second column address from a second command in the buffer. The device may determine whether the first command and the second command are hazardous commands based on comparing the first row address and the first column address from the first command with the second row address and the second column address from the second command.

Abstract: A data storage device whose controller is configured to apply a hash function to a logical address specified in a received host request to obtain a first portion of the corresponding physical address (e.g., the channel number or channel and die numbers). This feature of the controller enables the L2P table stored in the DRAM associated with the controller to have physical-address entries that contain therein only complementary second portions of the physical addresses, but not the first portions. Such shorter physical-address entries in the L2P table enable a corresponding beneficial reduction in the size of the DRAM and can further be leveraged to have optimized and aligned access to the L2P table in the DRAM.

Abstract: Methods, systems, and devices for post error correction code (ECC) registers for cache metadata are described. A device may read metadata from a memory array included in the device. The metadata may include information for operating a volatile memory as a cache for a non-volatile memory. The device may perform an ECC operation on the metadata based on reading the metadata from the memory array. After performing the ECC operation on the metadata, the device may write the metadata to a register that is coupled with the memory array. The device may then write the metadata from the register to the memory array.

Abstract: Methods, systems, and devices for memory wear management are described. A device may include an interface controller and a non-volatile memory. The interface controller may manage wear-leveling procedures for memory banks in the non-volatile memory. For example, the interface controller may select a row in a memory bank for the wear-leveling procedure. The interface controller may store data from the row in a buffer in the interface controller. The interface controller may then transfer the data to the non-volatile memory so that the non-volatile memory can write the data to a second row of the memory bank.

Abstract: Methods, systems, and devices for transaction management using metadata are described. In some examples, a memory device may include a volatile memory, and a non-volatile memory, which may have different access latencies. The memory device may receive from a host device a read command for data located at an address of the non-volatile memory. In response to the read command, the memory device and may determine whether the data is stored in the volatile memory. The memory device may then transmit, to the host device data and according to an expected latency, a set of data and an indication of whether the set of data was previously requested by the host device or unrequested by the host device. In some examples, the memory device may also transmit an identifier associated with the read command and a hash of the address.

Abstract: Methods, systems, and devices for compressed logical-to-physical mapping for a memory bypass for error detection and correction are described. A memory device may include error detection and correction circuitry for detecting and correcting errors in data that is read from a memory array of the memory device. To reduce read latencies, the memory device may include bypass circuitry that enables it to transmit the data to the host device before or during error detection. If the memory device determines that the data is erroneous, the memory device may transmit an alert to the host device concurrently with or after transmitting the data. The memory device may perform error correction on the data and store corrected data in a register. Based on receiving an alert, the host device may issue one or more additional read commands to re-read the data from the memory bank or read the corrected data from the register.

Abstract: A non-volatile storage system includes: a host and a storage device. The host includes a submission queue memory, a completion queue memory, and a read/write data memory, and the storage device includes: a controller configured to concurrently communicate with the read/write data memory and with at least one of the submission queue memory and the completion queue memory; and a memory device configured to communicate with the controller.

Abstract: Methods, systems, and devices for cache metadata management in a memory subsystem are described. The memory subsystem may include an interface controller coupled with a non-volatile memory and a volatile memory. The interface controller may use metadata, such as validity information and dirty information, to operate the volatile memory as cache. The interface controller may store the dirty information in the volatile memory and may store the validity information in an array in the interface controller.

Abstract: Methods, systems, and devices for transaction management based on metadata are described. A host device may transmit a read command to a memory device. Based on the read command, the host device may receive a set of data from the memory device. The host device may also receive metadata associated with the set of data. Based on the metadata, the host device may determine whether the set of data is the data requested by the read command, data requested by a previous read command, or data unrequested by the host device, or some combination. If the set of data is the data requested by the read command or a previous read command, the host device may process the set of data accordingly. If the set of data is data unrequested by the host device, the host device may discard the set of data and retransmit the read command.

Abstract: Methods, systems, and devices for hazard detection in a multi-memory device are described. A device may receive a first command that indicates a first bank address, a first row address, and a first column address. Based on the first bank address, the device may select a buffer for a hazard detection procedure that detects hazardous commands. The device may compare, as part of the hazard detection procedure, the first row address and the first column address from the first command with a second row address and a second column address from a second command in the buffer. The device may determine whether the first command and the second command are hazardous commands based on comparing the first row address and the first column address from the first command with the second row address and the second column address from the second command.

Abstract: Methods, systems, and devices for post error correction code (ECC) registers for cache metadata are described. A device may read metadata from a memory array included in the device. The metadata may include information for operating a volatile memory as a cache for a non-volatile memory. The device may perform an ECC operation on the metadata based on reading the metadata from the memory array. After performing the ECC operation on the metadata, the device may write the metadata to a register that is coupled with the memory array. The device may then write the metadata from the register to the memory array.

Abstract: Methods, systems, and devices for cache management in a memory subsystem are described. A device may determine to perform an eviction procedure for a bank of a volatile memory that operates as a cache for a non-volatile memory. The eviction procedure may save data from the bank of the volatile memory to the non-volatile memory. The device may determine an activity status for at least one buffer in a pool of buffers that are coupled with the volatile memory and the non-volatile memory. The device may select the at least one buffer in the pool of buffers for the eviction procedure for the bank of the volatile memory based at least in part on the activity status for that buffer.

Abstract: A non-volatile storage system includes: a host and a storage device. The host includes a submission queue memory, a completion queue memory, and a read/write data memory, and the storage device includes: a controller configured to concurrently communicate with the read/write data memory and with at least one of the submission queue memory and the completion queue memory; and a memory device configured to communicate with the controller.