Abstract: A data processing system and method for accessing a heterogeneous memory system including a processing unit are provided. The heterogeneous memory system includes a memory module and high bandwidth memory (HBM) including a processing-in-memory (PIM) circuit combined with a memory controller. The memory controller is configured to detect a data array required for an arithmetic operation from a memory module or the HBM by using a border index value when the arithmetic operation is performed by the PIM circuit of the HBM and generate a memory module command set and an HBM command set using physical address spaces respectively designated in the memory module and the HBM.

Abstract: According to one embodiment, a memory system includes a nonvolatile memory and a controller. The controller communicates with a host. The host includes a host memory and a circuit. The circuit accesses the host memory in a unit of first size. When an address designated as a first location of the host memory where data read from the nonvolatile memory is to be stored is not aligned with a boundary in the host memory defined in a unit of the first size, the controller transmits a first packet which has a size from the first location to the boundary and includes the read data to be stored from the first location, and transmits a second packet which has the first size and includes the read data to be stored from the boundary thereafter.

Abstract: Embodiments provide a read operation circuit, a semiconductor memory, and a read operation method. The read operation circuit includes: a data determination module configured to read read data from the memory bank, and determine, according to the number of bits of a data change between previous read data and current read data, whether to invert the current read data to output global bus data for transmission through a global bus and inversion flag data for transmission through an inversion flag signal line; a data receiving module configured to determine whether to invert the global bus data according to the inversion flag data to output cache data; a parallel-to-serial conversion circuit configured to perform parallel-to-serial conversion on the cache data to generate output data of a DQ port; and a data buffer module configured to determine an initial state of the global bus according to enable signal and current read data.

Abstract: A storage device includes; a nonvolatile storage including a first region and a second region, a storage controller controlling operation of the nonvolatile storage, and a buffer memory connected to the storage controller. The storage controller stores user data received from a host device in the second region, stores metadata associated with management of the user data and generated by a file system of the host device in the first region, loads the metadata from the first region to the buffer memory in response to address information for an index node (inode) associated with the metadata, and accesses the target data in the second region using the metadata loaded to the buffer memory.

Abstract: An in-place data recovery method and system include receiving a user request to restore a virtual machine to a version corresponding to a first point in time, identifying a first snapshot of the virtual machine based on the user request, generating a second snapshot of the virtual machine, identifying a second data block in the second snapshot that includes modified data derived from data content of a first data block in the first snapshot, generating reverse incremental backup data including the first data block, and restoring the virtual machine in-place based on the reverse incremental backup data.

Abstract: A processor having a functional slice architecture is divided into a plurality of functional units (“tiles”) organized into a plurality of slices. Each slice is configured to perform specific functions within the processor, which may include memory slices (MEM) for storing operand data, and arithmetic logic slices for performing operations on received operand data. The tiles of the processor are configured to stream operand data across a first dimension, and receive instructions across a second dimension orthogonal to the first dimension. The timing of data and instruction flows are configured such that corresponding data and instructions are received at each tile with a predetermined temporal relationship, allowing operand data to be transmitted between the slices of the processor without any accompanying metadata. Instead, each slice is able to determine what operations to perform on received data based upon the timing at which the data is received.

Abstract: A peripheral device may implement storage virtualization for non-volatile storage devices connected to the peripheral device. A host system connected to the peripheral device may host one or multiple virtual machines. The peripheral device may implement different virtual interfaces for the virtual machines or the host system that present a storage partition at a non-volatile storage device to the virtual machine or host system for storage. Access requests from the virtual machines or host system are directed to the respective virtual interface at the peripheral device. The peripheral device may perform data encryption or decryption, or may perform throttling of access requests. The peripheral device may generate and send physical access requests to perform the access requests received via the virtual interfaces to the non-volatile storage devices. Completion of the access requests may be indicated to the virtual machines via the virtual interfaces.

Abstract: A system startup method includes creating a first thread when a kernel driver in a kernel mode detects a first disk partition, reading, in the kernel mode, metadata of the first disk partition using the first thread, and writing the metadata of the first disk partition into a first page cache using the first thread. In the kernel mode, metadata of a disk partition is pre-cached into a page cache using the first thread, and in a subsequent process in a user mode, the metadata of the disk partition is directly read from the page cache. A storage area of the page cache is memory and has a higher read/write operation rate compared with a disk.

Abstract: A memory controller is configured with a plurality of processors to be operated in parallel so that overhead of firmware may be reduced. The memory controller includes a first processor and a second processor. The first processor is configured to generate a command corresponding to a request received from a host and to translate a logical address included in the request into a physical address of a memory device. The second processor is configured to operate on data to be output to the memory device or on data received from the memory device. If the request is received from the host when the second processor is in an idle state, the first processor may control the second processor to release the idle state of the second processor and to perform an operation of logging command information corresponding to the request.

Abstract: The memory sub-systems of the present disclosure selects, for memory scans, a memory block which has a highest page fill ratio. In one embodiment, the memory sub-system identifies a number of block stripes located on a logical unit (LU) identified by a logical unit number (LUN), where the LU is one of a plurality of LUs of a memory device. The sub-system determines a fill ratio for each of the plurality of block stripes. The sub-system selects, among the block stripes, a block stripe with a highest fill ratio. The sub-system identifies, from the selected block stripe, a memory block of the LU. The sub-system performs a memory scan operation on the memory block of the memory device.

Abstract: Methods, systems, and devices for parameter table protection for a memory system are described. Upon booting a memory system for a first time, the memory system or a host system may generate an error control code associated with parameter data stored to the memory system. The error control code may be stored to the memory system and may be configured to correct one or more errors in the parameter data upon subsequent boot sequences of the memory system. Accordingly, upon booting the memory system for a second or a subsequent time, the error control code may be used to identify and correct errors in the parameter data, which may reduce the quantity of copies of parameter data stored to the memory system and may prevent the memory system from experiencing a system crash.

Abstract: A disk device includes a volatile memory, a nonvolatile memory, and a controller. The controller is configured to receive, from a host, a key setting request that includes a cryptographic key, a key ID thereof, and tag information of the cryptographic key and generate generation information of the cryptographic key. The controller is also configured to store a first entry including the tag information, the cryptographic key, and the generation information associated with each other in the volatile memory, and store a second entry including the key ID and the generation information associated with each other in the nonvolatile memory.

Abstract: In some examples, a data node may provide a file system executed in a user space. The data node may invoke a device check process configured to perform a test for failure of a storage device associated with a target volume by sending a file system level request for obtaining metadata of data stored to the target volume. For instance, the target volume may be a virtual storage volume that represents storage capacity on a network storage. The user-space file system may receive the request for the metadata and may retrieve preconfigured metadata of the target volume that is stored locally in advance of the invoking of the device check process. The preconfigured metadata may be sent to the device check process in response to the request. For example, the preconfigured metadata may at least partially cause the target volume to pass the test.

Abstract: A primary storage array calculates signatures of chunks of production device data that are sent to a target device on a secondary storage array. The chunk signatures are sent to a signature device on the secondary storage array, where the chunk signatures are stored within the same LBA range on the signature device as their corresponding chunks are stored on the target device. Snaps of the target and signature device are created and associated as a snap pair. Later, the primary storage array calculates signatures of changed chunks of production device data that are sent to the target device. The changed chunk signatures are sent to the signature device. New snaps of the target and signature device are created and associated as a new snap pair. Chunk data is validated by calculating signatures of the chunks from the target device and comparing those signatures with the chunk signatures from the signature device.

Abstract: An illustrative method may include creating a volume object based on a request that includes a user token that indicates an identifier of a user associated with the request, the volume object corresponding to a physical or virtual volume of storage; creating an ownership object corresponding to the volume object, the creating comprising obtaining the identifier of the user from the user token and including the identifier in the ownership object; and including in the ownership object an access control list that is associated with an operation type, the access control list comprising identifiers of users and/or user groups who have permission to perform operations on the volume object.

Abstract: A trigger condition associated with latent read disturb in a memory device is detected. In response to detecting the trigger condition associated with latent read disturb, one or more blocks in the memory device that are impacted by the trigger condition are placed in a stable state to mitigate latent read disturb in the one or more blocks.

Abstract: According to one embodiment, a storage system performs a first allocation operation of allocating, for a first namespace, a plurality of first blocks included in the blocks of a nonvolatile memory. The storage system performs a read operation, a write operation or an erase operation on one of the first blocks in response to a command received from a host to read, write or erase the one first block, counts the total number of erase operations performed on the first blocks, and notifies the host of the counted number of erase operations in response to a command received from the host to obtain an erase count associated with the first namespace.

Abstract: Methods, systems, and devices for suspend operation with data transfer to a host system are described. A host system may transmit a read command to a memory system operating in a first mode of operation (e.g., a standard mode associated with a nominal power consumption) indicating for the memory system to transition to a second mode of operation (e.g., a suspend mode associated with a decreased power consumption). Here, the memory system may transmit an image of the memory system stored in volatile memory to the host system and transition the memory system to the second mode. Additionally, the host system may transmit, to the memory system operating in the second mode, a write command including the image and indicating for the memory system to transition to the first mode. Here, the memory system may write the image to the volatile memory and transition to the first mode.

Abstract: A data storage device stores data in non-volatile memory. In one approach, a method includes: storing software in a compressed format in a first mode (e.g., an SLC mode) in a non-volatile memory; exposing, while the software is stored in the first mode, the non-volatile memory to a temperature greater than a predetermined threshold; determining that the temperature of the non-volatile memory has fallen below the predetermined threshold; and in response to determining that the temperature of the non-volatile memory has fallen below the predetermined threshold: decompressing the stored software, and storing the decompressed software in a second mode (e.g., TLC mode) in the non-volatile memory. The second mode has a storage density higher than the first mode.

Abstract: Apparatus and methods are disclosed, including receiving an indication to selectively erase first data stored on a first page of a first subset of a group of multi-level memory cells of the storage system, each multi-level memory cell comprising multiple pages and providing, in response the indication to selectively erase the first data, at least one soft erase pulse to the first page of memory cells associated with the first data to induce distribution overlap across different bit levels of the first page of the group of multi-level memory cell.