Abstract: A method includes receiving a machine learning (ML) network model in high-level code; generating an internal representation (IR), the IR is mapped to components in a multi-processing tile device; determining whether a first processing tile with a first on-chip memory (OCM) has a same dimension for an input/output tensor data as a second processing tile with a second OCM performing a same primitive function based on the IR; allocating a same memory address range within the first and the second OCM for the same primitive function if the first processing tile has the same dimension for the input/output tensor data as the second processing tile for the same primitive function; linking the memory address range of the first OCM to the memory address range of the second OCM to form a grouped memory space within the first and the second OCM respectively; and compiling low-level instructions based on the linking.

Abstract: An object pose estimation system, an execution method thereof and a graphic user interface are provided. The execution method of the object pose estimation system includes the following steps. A feature extraction strategy of a pose estimation unit is determined by a feature extraction strategy neural network model according to a scene point cloud. According to the feature extraction strategy, a model feature is extracted from a 3D model of an object and a scene feature is extracted from the scene point cloud by the pose estimation unit. The model feature is compared with the scene feature by the pose estimation unit to obtain an estimated pose of the object.

Abstract: Semiconductor structures and methods are provided. In one embodiment, a method of the present disclosure includes forming a plurality of semiconductor fins over a substrate, after the forming of the plurality of semiconductor fins, removing an outer semiconductor fin of the plurality of semiconductor fins, and forming a gate structure over the plurality of semiconductor fins. The plurality of semiconductor fins include more than 3 semiconductor fins and the removing recesses a portion of the substrate directly under the outer semiconductor fin.

Abstract: A method of manufacturing a semiconductor device includes forming a plurality of fin structures extending in a first direction over a semiconductor substrate. Each fin structure includes a first region proximate to the semiconductor substrate and a second region distal to the semiconductor substrate. An electrically conductive layer is formed between the first regions of a first adjacent pair of fin structures. A gate electrode structure is formed extending in a second direction substantially perpendicular to the first direction over the fin structure second region, and a metallization layer including at least one conductive line is formed over the gate electrode structure.

Abstract: The present invention relates to an energy consumption regulation method. The method includes executing the following steps by a processor: continuously collecting and learning a plurality of first operation data about an electronic device, thereby automatically distinguishing an operation of the electronic device into a plurality of operation periods; automatically configuring a different plurality of energy configurations for the electronic device according to the different plurality of operation periods to regulate a maximum power of the processor, an operation mode of a backlight unit, a charging mode of a battery, and a peripheral energy management mode; and continuously collecting and learning a plurality of second operation data about the electronic device and autonomously updating and adjusting the plurality of operation periods and the plurality of energy configurations based on the plurality of second operation data.

Abstract: An optical device includes a range finding module. The range finding module includes a first light condenser unit, a light emitting unit and a light receiving unit. The first light condenser unit defines an optical axis and a hole disposed along the optical axis. The first light condenser unit, the light emitting unit and the light receiving unit are sequentially arranged along the optical axis. The light is emitted by the light emitting unit, passes through the hole, reaches an object, is reflected by the object, is converged by the first light condenser unit and is received by the light receiving unit to generate an electrical signal.

Abstract: A method performed by an 8-TX UE for full power transmissions is provided. The method transmits an IE for reporting capability information. The method receives a PUSCH configuration for configuring the 8-TX UE with a CB PUSCH transmission, a full power mode 1, and one of a partial coherent codebook subset with two antenna groups, a partial coherent codebook subset with four antenna groups, and a non-coherent codebook subset. The method receives DCI including an SRI and an additional TPMI for indicating an 8-port SRS resource. The method then performs the CB PUSCH transmission using the full power mode 1. The IE includes a field indicating support for the full power mode 1 and a field indicating support for one of the partial coherent codebook subset with two antenna groups, the partial coherent codebook subset with four antenna groups, and the non-coherent codebook subset.

Abstract: The invention provides a metal oxide semiconductor (MOS) capacitor structure, which includes a counter-doping region in the channel region directly below the gate. Between the deep ion well and the counter-doping region is a semiconductor region. The doping concentration of the semiconductor region is lower than that of the deep ion well. The P-type well ion implantation processes in the active region of the device can be omitted, so the production cost is lower, and the dosage of the counter-doping region can be reduced, which improves the time-dependent dielectric collapse (TDDB) issue.

Abstract: A method performed by a user equipment (UE) for network energy saving (NES) is provided. The method receives, from a base station (BS), a channel state information (CSI) report configuration including multiple sub-configurations associated with a triggering state, each of the multiple sub-configurations indicating a list of channel state information (CSI)-reference signal (RS) identifiers (IDs), first information for a spatial domain adaptation, and second information for a power domain adaptation. The method receives multiple CSI-RSs based on the list of CSI-RS IDs, the first information, and the second information indicated by each sub-configuration in a subset of the multiple sub-configurations. The method generates a CSI report by measuring the multiple CSI-RSs. The method then transmits, to the BS, the CSI report. The triggering state is associated with aperiodic CSI reporting or semi-persistent CSI reporting.

Abstract: A method for a User Equipment (UE) for enhancement in Network Energy Saving (NES) is provided. The method receives, from a Base Station (BS), a Radio Resource Control (RRC) message including a trigger state list and a Channel State Information (CSI) report configuration that includes one or more CSI report sub-configurations. The method receives, from the BS, a Downlink Control Information (DCI) format indicating a trigger state in the trigger state list. The trigger state is associated with the CSI report configuration and indicates a subset of CSI report sub-configurations in the one or more CSI report sub-configurations. The method then applies the subset of CSI report sub-configurations to perform a CSI reporting procedure on a Physical Uplink Shared Channel (PUSCH).

Abstract: A system for decentralized identification of a file access permission includes a front-end program and an access control server. The front-end program includes at least one credential. The front-end program extracts attribute data of a corresponding access condition and a definition file identifier (ID) of the credential from the credential based on a check rule message related to a specified file, to form a multi-credential proof file. The access control server obtains an issuer decentralized identifier (DID) and a first public key corresponding to the multi-credential proof file by querying a blockchain for all of the credential definition files corresponding to the multi-credential proof file, confirms that each piece of the attribute data in the multi-credential proof file is correct, and all of the attribute data in the multi-credential proof file meets the access condition, and then grants the front-end program an access permission corresponding to the specified file.

Abstract: A method for performing an RA procedure for LTM is provided. The method receives a PDCCH order for initiating the RA procedure for an LTM candidate cell. The PDCCH order indicates that an index of a preamble is not “0.” The PDCCH order includes a first DCI field indicating that a transmission of the preamble is an initial transmission or a re-transmission and a second DCI field indicating the LTM candidate cell. The method increments a power ramping counter by 1 when the first DCI field indicates the re-transmission and the PDCCH order indicates an SSB for the re-transmission of the preamble. The SSB is the same as an SSB used in a previous transmission of the preamble. The method sets, based on the incremented power ramping counter, a power parameter. The method re-transmits, to the LTM candidate cell, the preamble based on at least the set power parameter.

Abstract: A method performed by a User Equipment (UE) for Network Energy Saving (NES) is provided. The method receives, from a Base Station (BS), a Radio Resource Control (RRC) message including a Transmission Configuration Indication (TCI) list. The method stores the TCI list in the UE. The method receives, from the BS, a Downlink Control Information (DCI) format. The method then updates the TCI list stored in the UE based on the DCI format.

Abstract: A method of slicing wafers from a monocrystalline semiconductor ingot includes attaching a circumferential edge of the ingot to a bond beam and positioning sacrificial disks adjacent longitudinal end faces of the ingot. One sacrificial disk is positioned adjacent each of the longitudinal end faces. The method also includes connecting the bond beam to a wire saw that includes a wire web and performing a slicing operation on the ingot by operating the wire saw to drive the wire web and move the bond beam and the ingot in a movement direction towards the wire web to slice the wafers from the ingot.

Abstract: A method for performing codebook-based Physical Uplink Shared Channel (PUSCH) transmission by a User Equipment (UE) is provided. The method receives a Radio Resource Control (RRC) message including a PUSCH configuration indicating a codebook-based PUSCH transmission. The method receives a Downlink Control Information (DCI) format including a first field and a second field. The first field indicates an 8-port Sounding Reference Signal (SRS) resource. The second field indicates an index associated with a Transmission Precoding Matrix Indicator (TPMI) and the number of transmission layers. The method determines the number of codewords to be one or two based on the number of transmission layers. The method determines a codeword-to-layer mapping for spatial multiplexing based on the number of transmission layers and the number of codewords. The method determines a precoding matrix based on the index and the codeword-to-layer mapping.

Abstract: A system for slicing wafers from a monocrystalline semiconductor ingot includes a wire saw, a bond beam, the monocrystalline semiconductor ingot, and two sacrificial disks. The wire saw includes a wire web and wire guides operable to drive the wire web during a slicing operation. The bond beam is connected to the wire saw. The wire saw is operable to move the bond beam in a movement direction towards the wire web during the slicing operation to slice the wafers from the ingot. The ingot includes longitudinal end faces and a circumferential edge extending between the longitudinal end faces. The ingot is attached to the bond beam along the circumferential edge. One sacrificial disk is positioned adjacent each of the longitudinal end faces of the ingot to inhibit uncontrolled breakage of the wafers during the slicing operation.

Abstract: A system and method for splitting a machine learning (ML) graph is disclosed. The system includes a compiler configured to receive an ML model. The compiler generates a graph associated with the ML model, wherein the graph is an internal representation of the ML model. The graph is partitioned into a first subgraph and a second subgraph. The first subgraph is associated with an ML hardware, an ML emulator, or a combination thereof, and the second subgraph is associated with a processor different from the ML hardware. A set of low-level instructions associated with the first subgraph is generated. One or more resources in the ML hardware is identified to execute the set of low-level instructions associated with the first subgraph.

Abstract: A satellite communication system and a method for managing radio resource of a non-terrestrial network are provided. The method includes: transmitting, by a first satellite, a first resource scheduling assignment when leaving a service area of the non-terrestrial network; receiving, by a second satellite, a second resource scheduling assignment corresponding to the first scheduling assignment when entering the service area; and accessing, by the second satellite, the radio resource according to the second resource scheduling assignment.

Abstract: A new approach is proposed to support correlating high-level code with low-level instructions of an application running on a hardware. A compiler that compiles a high-level function in the high-level code of the application into a set of low-level instructions to be executed on the hardware is configured to utilize one or more reserved fields of the set of low-level instructions to incorporate one or more IDs and an actionable item. The IDs are mapped to the high-level function, wherein such mapping is programmable by the compiler. Based on the mapped IDs and the actionable item incorporated in the set of the low-level instructions, the runtime performance of the application on the hardware can be monitored and profiled and issues related to the high-level code of the application can be identified for debugging purposes.

Abstract: A distributed data processing system and a distributed data processing method are provided. The distributed data processing system includes a computing device and at least one additional computing device.