Abstract: A data processing method includes determining a current data operation state of a follower storage node based on receiving a data read request of a terminal, the current data operation state comprising a read-only (RO) state and a read-write (RW) state; determining a reference log index of a leader storage node based on the current data operation state, the reference log index being a recently stored log index based on both the leader storage node and the follower storage node being in the RW state; acquiring a synchronized log index stored in the follower storage node, the synchronized log index being obtained based on synchronizing a log index stored in the leader storage node; acquiring data requested by the data read request from the follower storage node based on the synchronized log index covering the reference log index; and providing the data to the terminal.

Abstract: A storage device organizes data on different storage media based on access history and user behavior to accelerate data retrieval. The storage device includes a first storage media and a second storage media to store the data. The second storage media may be accessed faster than the first storage media. A controller on the storage device executes a first data retrieval to identify hot blocks in the first storage media based on the access history. The controller also executes a second data retrieval to identify the hot blocks in the first storage media based on a semantic search. The controller may prefetch the hot blocks from the first storage media, cache the prefetched hot blocks in the second storage media and/or move the hot blocks from a slower storage medium to a faster storage medium.

Abstract: A storage device organizes data on different storage media based on access history and user behavior to accelerate data retrieval. The storage device includes a first storage media and a second storage media to store the data. The second storage media may be accessed faster than the first storage media. A controller on the storage device executes a first data retrieval to identify hot blocks in the first storage media based on the access history. The controller also executes a second data retrieval to identify the hot blocks in the first storage media based on a semantic search. The controller may prefetch the hot blocks from the first storage media, cache the prefetched hot blocks in the second storage media and/or move the hot blocks from a slower storage medium to a faster storage medium.

Abstract: Three-dimensional (3D) semiconductor devices and fabricating methods are provided. In some implementations, a 3D semiconductor device includes: an array of vertical transistors each comprising a semiconductor body extending in a vertical direction; a plurality of word lines each extending in a first lateral direction, wherein each word line is shared by a row of the vertical transistors arranged along the first lateral direction; and a plurality of bit lines each extending in a second lateral direction perpendicular to the first lateral direction; wherein the semiconductor bodies are further arranged along a third lateral direction different from the first lateral direction and the second lateral direction.

Abstract: In certain aspects, a memory device includes an array of memory cells, and a plurality of peripheral circuits coupled to the array of memory cells and configured to control the array of memory cells. A first peripheral circuit of the plurality of peripheral circuits includes a first three-dimensional (3D) transistor. The first 3D transistor includes a 3D semiconductor body, and a gate structure in contact with a plurality of sides of the 3D semiconductor body. The gate structure includes a gate dielectric and a gate electrode. The gate electrode includes a metal. The gate dielectric has a thickness between 1.8 nm and 10 nm.

Abstract: Technology for NAND in-memory compute. NAND memory cells are organized into basic compute engines (CE). A basic CE contains a group of NAND memory cells in the same plane in the memory system. A basic CE may be associated with a set of bit lines in the plane. The memory system may map vectors of leaf nodes of one or more trees to basic CEs and program the vectors of the leaf nodes into basic compute engines in accordance with the mapping. The memory system perform an in-memory vector-vector multiplication in parallel between an input vector and each of the vectors of one or more of the leaf nodes in one or more of the basic compute engines.

Abstract: Example channel circuits, data storage devices, and methods for using a trained neural network to estimate the noise mixture in a read signal are described. Samples are determined from a digital read signal, such as the read signal from the non-volatile storage medium of a data storage device. The samples are processed through one or more instances of a neural network comprised of trained coefficients and outputting a set of estimate values for a noise mixture of the read signal. The set of estimate values may then be used to adjust parameters of the read channel for processing the read signal to detect and decode data.

Abstract: A manganese steel consisting essentially of the following chemical composition in mass percentages: C: 0.05%-0.10%, Mn: 4.1%-4.7%, Si: 0.15%-0.4%, P?0.010%, S?0.003%, Ti: 0.01%-0.05%, Ni+Cr+Mo?0.6%, and the balance of Fe and unavoidable impurities, the steel having a tensile strength between about 785 MPa to about 856 MPa.

Abstract: Examples of the present disclosure include an energy-storage system having a plurality of energy units, each energy unit including a battery module and a power converter coupled to a respective battery module; and at least one controller configured to (a) receive one or more parameter values from each battery module, each parameter value indicating a respective battery module state, (b) determine a difference between a largest parameter value from a first battery module and a smallest parameter value from a second battery module, (c) determine whether the difference is greater than a threshold difference, and (d) control at least one power converter to transfer energy between the first battery module and the second battery module through a lowest number of power converters in response to determining that the difference is greater than the threshold difference.

Abstract: Disclosed is an ensemble coupled assimilation system for numerical prediction and the method thereof. The ensemble coupled assimilation system includes a module for integrating coupled models and managing collaborative ensemble operation, an assimilation algorithms integration module, a module for configuring ensemble coupled assimilation experiment, and an online interaction module for ensemble coupled assimilation.

Abstract: A method includes: obtaining a media access control MAC address of a terminal device by using a routing device, where the terminal device is connected to the routing device; sending a packet to the terminal device based on the MAC address; receiving, from the terminal device, a response to the packet; and obtaining channel state information CSI between a sensing device and the terminal device based on the response, to perform sensing.

Abstract: To accelerate search speeds for approximate nearest neighbor searches of vector databases, compute-in-memory techniques using NAND memory structures are introduced. For each element of the database, a kernel of its M nearest neighbors is determined. For each vector of the database, both the vector and its kernel are programmed in the arrays of a NAND memory based accelerator card, so that the vectors will be written into the memory arrays both as themselves and also in kernels of vectors for which they are a nearest neighbor. Metadata, associating the locations of the kernel members with the correspond vector is also stored in the memory system. After determining the input's nearest neighbor at one level of search, the metadata is then used to locate that nearest neighbor's nearest neighbors and their distances to the input vector are then computed in parallel in a compute-in-memory vector-vector dot product multiplication.

Abstract: The present disclosure is directed to an NFC wireless charging system for electronic devices. The system includes a power transmitting device having a charging transmitter, and a power receiving device having a charging receiver. Each of the charging transmitter and the charging receiver implements a coil antenna design in which the coil antenna charging the power receiver is inserted into the coil antenna of the charging transmitter, or vice versa.

Abstract: A semiconductor device includes a first channel structure extending in a first direction in a stack structure, a conductive layer extending in a second direction perpendicular to the first direction and surrounding at least partial of the first channel structure, a first isolation structure extending in the first direction and the second direction, and a second isolation structure extending in the first direction and the second direction. The first channel structure and the conductive layer are disposed between the first isolation structure and the second isolation structure, and a first distance between the first channel structure and the first isolation structure is less than a second distance between the first channel structure and the second isolation structure.

Abstract: A Data Storage Device (DSD) includes a first memory storing reference files used to derive vector embeddings in a vector database. A query vector embedding is received from a host and one or more vector embeddings similar to the query vector embedding are identified in the vector database. One or more reference files from which the one or more vector embeddings were derived are identified and stored in a second memory for faster access. In another aspect, a query vector embedding is received by a host that identifies one or more vector embeddings that are similar to the query vector embedding and retrieves the one or more vector embeddings from a DSD to provide to an Artificial Intelligence (AI) model. One or more reference files are identified from which the one or more vector embeddings were derived and are prefetched from the DSD for storage in a host memory.

Abstract: Semiconductor devices, methods for forming such semiconductor devices, and systems including such semiconductor devices are provided. In one aspect, a semiconductor device includes a first semiconductor structure that includes a stack structure and a transistor. The stack structure includes first conductive layers and second conductive layers that are stacked alternately in a first direction. The transistor is disposed on one side of the stack structure and connected with one of the first conductive layers. The first semiconductor structure further includes a first connection structure extending through the stack structure in the first direction and connected to the second conductive layers.

Abstract: A device for storing a sheet pile in a converting machine (10) comprises a transfer mechanism (28) for delivering sheets (22) into a piling area (26) of the device (22), and a sampling plate (44) which is adjustable between a sampling position and an access position, wherein the top of the piling area (26) is covered by the sampling plate (44) when the sampling plate (44) is in the sampling position and the top of the piling area (26) is free when the sampling plate (44) is in the access position.

Abstract: A method, an apparatus, a system, a device, and a storage medium for implementing terminal verification are provided. In this method, a transmission feature of a terminal is reconstructed to verify the terminal. For example, if a difference between the transmission feature obtained through reconstruction and the transmission feature of the terminal is relatively large, it indicates that the transmission feature of the terminal is abnormal, and the terminal is an abnormal terminal. In this case, it is determined that the verification of the terminal fails.

Abstract: Devices and methods to implement semantic searching on SSD through a computational SSD system that distributes computing to each NAND flash die of the SSD while the SSD controller handles the results aggregation with new on-die computation logic circuits to provide on device file semantic search are disclosed herein. The computational SSD system can read file feature vectors from multiple dies to the SSD controller, and if needed, these feature vectors may be buffered in DRAM and controller handles distance computing. Local, on-die AI/ML processing units may perform, for example, computation and comparison operations and pass the processing scores and results to the SSD controller. The SSD controller aggregates results from all dies and returns the result to the host. The feature vector store size, circuitry and number of on-die AI/ML processing units may be configured as needed to adapt to different tasks, system constraints, and/or feature vector sizes.

Abstract: A Data Storage Device (DSD) includes a first memory storing reference files used to derive vector embeddings in a vector database. A query vector embedding is received from a host and one or more vector embeddings similar to the query vector embedding are identified in the vector database. One or more reference files from which the one or more vector embeddings were derived are identified and stored in a second memory for faster access. In another aspect, a query vector embedding is received by a host that identifies one or more vector embeddings that are similar to the query vector embedding and retrieves the one or more vector embeddings from a DSD to provide to an Artificial Intelligence (AI) model. One or more reference files are identified from which the one or more vector embeddings were derived and are prefetched from the DSD for storage in a host memory.