A memory allocation device on an originating node requests an allocation of memory from a remote node. In response, the memory allocation device on the remote node returns a global system address that can be used to access the remote allocation from the originating node. Concurrent with the memory allocation device assigning (associating) a local (to its node) physical address to be used to access the remote allocation, the remote node allocates local physical memory to fulfill the remote allocation request. In this manner, the remote node has already completed the overhead operations associated with the remote allocation requested by the time the remote allocation is accessed by the originating node.

A non-volatile memory device includes arrays of non-volatile memory cells that are configured to the store weights for a recurrent neural network (RNN) inference engine with a gated recurrent unit (GRU) cell. A set three non-volatile memory arrays, such as formed of storage class memory, store a corresponding three sets of weights and are used to perform compute-in-memory inferencing. The hidden state of a previous iteration and an external input are applied to the weights of the first and the of second of the arrays, with the output of the first array used to generate an input to the third array, which also receives the external input. The hidden state of the current generation is generated from the outputs of the second and third arrays.

Systems, apparatuses, and methods related to bit string operations in memory are described. The bit string operations may be performed within a memory array without transferring the bit strings or intermediate results of the operations to circuitry external to the memory array. For instance, sensing circuitry that can include a sense amplifier and a compute component can be coupled to a memory array. A controller can be coupled to the sensing circuitry and can be configured to cause one or more bit strings that are formatted according to a universal number format or a posit format to be transferred from the memory array to the sensing circuitry. The sensing circuitry can perform an arithmetic operation, a logical operation, or both using the one or more bit strings.

Aspects of the present disclosure are directed to performing varying frequency memory sub-system background scans using either or both a timer and an I/O event limit. This can be accomplished by identifying a background scan trigger event from one of multiple possible types of background scan trigger events, such as a timer expiration or reaching an event count limit. In response to the background scan trigger event, a background scan can be initiated on a memory portion. The background scan can produce results, such as CDF-based data. When a metric based on the results exceeds a background scan limit, a refresh relocation can be performed and logged. A metric can be generated based on the CDF-based data, obtained error recovery depth data, or refresh relocation event data. When the metric is above or below corresponding background scan thresholds, a background scan frequency can be adjusted.

In a complex system including; one or more storage systems including a cache and a storage controller; and one or more storage boxes including a storage medium, the storage box generates redundant data from write data received from a server, and writes the write data and the redundant data to the storage medium. The storage box transmits the write data to the storage system when it is difficult to generate the redundant data or it is difficult to write the write data and the redundant data to the storage medium. The storage system stores the received write data in the cache.

Systems and methods are described for resource-efficient memory deduplication and write-protection. In an example, a method includes receiving, by a computing device having a processor, a request to assess deduplication for a plurality of candidate files. The computing device may perform one or more iterative steps for deduplication. The iterative steps may include: receiving, from the plurality of candidate files, a candidate file that is not write-protected; determining, based on a predetermined Bernoulli distribution, a decision to write-protect the candidate file; rendering the candidate file as a write-protected candidate file; determining, based on a review of other candidate files from the plurality of candidate files, that the write-protected candidate file can be deduplicated; and deduplicating the write-protected candidate file.

Examples of the present disclosure generally relate to integrated circuits, such as a system-on-chip (SoC), that include a memory subsystem. In some examples, an integrated circuit includes a first master circuit in a first power domain on a chip; a second master circuit in a second power domain on the chip; and a first memory controller in a third power domain on the chip. The first master circuit and the second master circuit each are configured to access memory via the first memory controller. The first power domain and the second power domain each are separate and independent from the third power domain.

Various examples may include an apparatus including a memory to store ingressing data or egressing data, a timer to generate a timing signal responsive to a user-configurable time interval, and a circuit to move the ingressing data or the egressing data from the memory at least partially responsive to the timing signal generated by the timer. Various examples may include a method including receiving a data packet at a network-facing interface, writing data of the data packet into a memory, receiving a timing signal, and responsive to the timing signal, providing the data from the memory at a device-facing interface. Various examples may include a method including receiving data at a device-facing interface, writing the data to a memory, receiving a timing signal, and responsive to the timing signal, providing a data packet including the data at a network-facing interface. Related devices, systems and methods are also disclosed.

A method for managing a memory buffer, a memory control circuit unit, and a memory storage apparatus are provided. The method includes the following steps. Multiple consecutive first commands are received from a host system. A command ratio of read command among the first commands is calculated. The memory storage apparatus is being configured in a first mode or a second mode according to the command ratio and a ratio threshold. A first buffer is configured in a buffer memory to temporarily store a logical-to-physical address mapping table in response to the memory storage device being configured in the first mode, in which the first buffer has a first capacity. A second buffer is configured in the buffer memory in response to the memory storage device being configured in the second mode, in which the second buffer has a second capacity, which is greater than the first capacity.

A memory system includes multiple dice having multiple planes. A processing device is coupled to the dice and performs controller operations including receiving a status indicator signal comprising a pulse that is asserted by one or more planes of the multiple dice. In response to receiving the pulse, the processing device performs at least one of: a first status check of dice operations being performed by the multiple dice at an expiration of a polling delay period; or a second status check of the dice operations in response to detecting the pulse being deasserted. The processing device terminates performances of status checks while the status indicator signal remains deasserted.