Abstract: Apparatus and methods are provided for thermal throttling for UE configured with multi-panel transceiving on FR2. In one novel aspect, the UE prioritizes throttling actions based on signal qualities of each transceiving panel. In one embodiment, the switching to the target panel from the active panel is selected as the highest priority throttling action when the signal quality of the target panel is similar to the active panel. In another embodiment, the UE further determines if the quality of the target panel is sufficient to support mmW transceiving before switching to the target panel. In one embodiment, the UE reduces one or more antennae of an active panel when the signal quality difference between the active panel and the target panel is bigger than a predefined gap threshold.

Abstract: The present disclosure describes a method to form a backside power rail (BPR) semiconductor device with an air gap. The method includes forming a fin structure on a first side of a substrate, forming a source/drain (S/D) region adjacent to the fin structure, forming a first S/D contact structure on the first side of the substrate and in contact with the S/D region, and forming a capping structure on the first S/D contact structure. The method further includes removing a portion of the first S/D contact structure through the capping structure to form an air gap and forming a second S/D contact structure on a second side of the substrate and in contact with the S/D region. The second side is opposite to the first side.

Abstract: The present disclosure is directed to methods for the fabrication of buried layers in gate-all-around (GAA) transistor structures to suppress junction leakage. In some embodiments, the method includes forming a doped epitaxial layer on a substrate, forming a stack of alternating first and second nano-sheet layers on the epitaxial layer, and patterning the stack and the epitaxial layer to form a fin structure. The method includes forming a sacrificial gate structure on the fin structure, removing portions of the fin structure not covered by the sacrificial gate structure, and etching portions of the first nano-sheet layers. Additionally, the method includes forming spacer structures on the etched portions of the first nano-sheet layers and forming source/drain (S/D) epitaxial structures on the epitaxial layer abutting the second nano-sheet layers.

Abstract: The present disclosure provides a method for fabricating an image sensor. The method includes the following operations. A cavity is formed at a first surface of a substrate. A germanium layer is formed in the cavity. A first heavily doped region is formed in the germanium layer by an implantation operation. A second heavily doped region is formed at a position proximal to a top surface of the germanium layer, wherein the second heavily doped region is laterally surrounded by the first heavily doped region from a top view perspective. An interconnect structure is formed over the germanium layer.

Abstract: Structures with doping free connections and methods of fabrication are provided. An exemplary structure includes a substrate; a first region of a first conductivity type formed in the substrate; an overlying layer located over the substrate; a well region of a second conductivity type formed in the overlying layer; a conductive plug laterally adjacent to the well region and extending through the overlying layer to electrically contact with the first region; and a passivation layer located between the conductive plug and the well region.

Abstract: A method includes: obtaining impact values for characteristic conditions; selecting training data subsets respectively from training data sets according to the impact values; obtaining a candidate model and an evaluation value based on the training data subsets; supplementing the training data subsets according to the impact values; obtaining another candidate model and another evaluation value based on training data subsets thus supplemented; repeating the step of supplementing the training data subset, and the step of obtaining another candidate model and another evaluation value based on the training data subsets thus supplemented; and selecting one of the candidate models as a prediction model based on the evaluation values.

Abstract: A waterproof click pad device includes a click pad, a frame and a waterproof unit. The frame surrounds the click pad and surrounds an axis passing through the click pad. The waterproof unit is transverse to the axis and is in sheet form. The waterproof unit includes a frame adhesive member surrounding the axis and adhered to the frame, a first non-adhesive member surrounding the axis, connected to an inner periphery of the frame adhesive member and spaced apart from and located above the frame, a second non-adhesive member surrounding the axis, connected to an inner periphery of the first non-adhesive member and spaced apart from and located above the click pad and the frame, and an plate adhesive member connected to an inner periphery of the second non-adhesive member and adhered to the click pad.

Abstract: The present disclosure relates to an integrated chip. The integrated chip includes a substrate having a first semiconductor material. A second semiconductor material is disposed on the first semiconductor material and a passivation layer is disposed on the second semiconductor material. A first doped region and a second doped region extend through the passivation layer and into the second semiconductor material. A silicide is arranged within the passivation layer and along tops of the first doped region and the second doped region.

Abstract: A gamma tap circuit includes: (i) a first gamma division circuit configured to generate a first gamma tap voltage by performing voltage division of an upper gamma tap voltage and a lower gamma tap voltage, in-sync with a first clock signal CK1 and a first complementary clock signal CK1b, which is 180° out-of-phase relative to CK1, (ii) a second gamma division circuit configured to generate a second gamma tap voltage by performing voltage division of the upper gamma tap voltage and the first gamma tap voltage, in-sync with a second clock signal CK2 and a second complementary clock signal CK2b, which is 180° out-of-phase relative to CK2, and (iii) a third gamma division circuit configured to generate a third gamma tap voltage by performing voltage division of the first gamma tap voltage and the lower gamma tap voltage, in response to CK2 and CK2b, which have a lower frequency relative to CK1 and CK1b.

Abstract: A liquid storage tank includes a housing, a piston located in the housing, a cover, an elastic element, and an outlet pipe. The cover is attached to the housing and has a support post extending toward the piston. The piston, the housing, and the cover define a tank chamber. The tank chamber is filled with cooling liquid. The elastic element is connected with the tank hosing and the piston. The elastic element is free from contact with the cooling liquid. The outlet pipe communicates with the tank chamber. An extension direction of an opening of the outlet pipe is not parallel to a direction of movement of the elastic element. When the cooling liquid is decreased, the piston compressed the tank chamber such that the elastic element is released. The tank chamber is continuously compressed by pairing the elastic element and the piston.

Abstract: A network quality measurement method and system are provided. In the method, a movement path and a movement speed of a vehicle device are determined according to a size of a space and an endurance time of the vehicle device, and the vehicle device is controlled to move on the movement path at the movement speed. During a movement of the vehicle device, a network quality in the space is measured according to a measurement frequency to generate network quality data. Whether the network quality in the space is changed is determined according to the network quality data. Whether there is an obstacle around the vehicle device is detected. When it is determined that the network quality in the space is changed or the obstacle is detected around the vehicle device, at least one of the movement path, the movement speed, and the measurement frequency is adjusted.

Abstract: A soft-switching power converter includes a main switch, an energy-releasing switch, and an inductive coupled unit. The main switch is a controllable switch. The energy-releasing switch is coupled to the main switch. The inductive coupled unit is coupled to the main switch and the energy-releasing switch. The inductive coupled unit includes a first inductance, a second inductance coupled to the first inductance, and an auxiliary switch unit. The auxiliary switch unit is coupled to the second inductance to form a closed loop. The main switch and the energy-releasing switch are alternately turned on and turned off. The auxiliary switch unit is controlled to start turning on before the main switch is turned on so as to provide at least one current path.

Abstract: A method performed by one or more computers for obtaining an optimized algorithm that (i) is functionally equivalent to a target algorithm and (ii) optimizes one or more target properties when executed on a target set of one or more hardware devices. The method includes: initializing a target tensor representing the target algorithm; generating, using a neural network having a plurality of network parameters, a tensor decomposition of the target tensor that parametrizes a candidate algorithm; generating target property values for each of the target properties when executing the candidate algorithm on the target set of hardware devices; determining a benchmarking score for the tensor decomposition based on the target property values of the candidate algorithm; generating a training example from the tensor decomposition and the benchmarking score; and storing, in a training data store, the training example for use in updating the network parameters of the neural network.

Abstract: A bonding structure that may be used to form 3D-IC devices is formed using first oblong bonding pads on a first substrate and second oblong bonding pads one a second substrate. The first and second oblong bonding pads are laid crosswise, and the bond is formed. Viewed in a first cross-section, the first bonding pad is wider than the second bonding pad. Viewed in a second cross-section at a right angle to the first, the second bonding pad is wider than the first bonding pad. Making the bonding pads oblong and angling them relative to one another reduces variations in bonding area due to shifts in alignment between the first substrate and the second substrate. The oblong shape in a suitable orientation may also be used to reduce capacitive coupling between one of the bonding pads and nearby wires.

Abstract: A probe card includes a flexible inorganic material layer, a metal micro structure, and a circuit board. The flexible inorganic material layer has a first surface and a second surface opposite to each other. The metal micro structure is disposed on the first surface. The circuit board is disposed on the second surface, and the circuit board is electrically connected to the metal micro structure. The test signal is adapted to be conducted to the circuit board through the flexible inorganic material layer.

Abstract: An electronic device including a hinge module, a first body, a second body, and a flexible display assembled to the first body and the second body is provided. Each of the first body and the second body is pivoted and slidably connected to the hinge module, and a cover of the hinge module is exposed out of the first body and the second body. The first body and the second body are rotated relatively via the hinge module to bend or flatten the flexible display, when the flexible display is bending from a flat state, a bending portion of the flexible display leans against the cover and pushes the cover away from the first body and the second body.

Abstract: A container feeding device includes a casing and first and second latch members. The casing defines a lower retaining space for receiving a plurality of containers that are stacked on one another. The first latch member is operable to enter the lower retaining space for supporting a bottommost container, or leave the lower retaining space to release the bottommost container. The second latch member enters the lower retaining space to support a second bottommost container when the bottommost container is released by the first latch member.

Abstract: A tool set includes a tool holder, a tool and a tool rack. The tool has a groove unit. The tool holder has a latch unit that engages the groove unit. The tool rack includes a rack body and a blocking member. When the tool holder is moved away from the rack body after the tool is moved into the rack body by the tool holder and after the blocking member moves to a blocking position, the tool is blocked by the blocking member so that the latch unit is separated from the groove unit and that the tool holder is separated from the tool.

Abstract: A method of performing automatic tuning on a deep learning model includes: utilizing an instruction-based learned cost model to estimate a first type of operational performance metrics based on a tuned configuration of layer fusion and tensor tiling; utilizing statistical data gathered during a compilation process of the deep learning model to determine a second type of operational performance metrics based on the tuned configuration of layer fusion and tensor tiling; performing an auto-tuning process to obtain a plurality of optimal configurations based on the first type of operational performance metrics and the second type of operational performance metrics; and configure the deep learning model according to one of the plurality of optimal configurations.

Abstract: A bridge control chip includes a first interface, a second interface, and a processor, wherein the first interface is coupled to a host device, the second interface is coupled to a memory device, and the memory device is a flash memory device. The processor is arranged to execute commands in a queue in sequence, to transmit the commands in the queue to the memory device through the second interface in sequence, wherein when the processor receives one or more received commands from the host device, the processor sorts the one or more received commands and commands which are currently and temporarily stored in the queue according to a distance between a logical address of each of the one or more received commands and a logical address of a current command in the queue that is currently executed by the processor.