Abstract: A data processing system is for uploading an upload data, wherein the upload data comprises a data content and a table name. The data processing system comprises a relational database; a NoSQL database; a data processing unit, coupled to the relational database and the NoSQL database, configured to execute a program code; and a storage unit, coupled to the data processing unit, configured to store the program code to instruct the data processing unit to execute a data processing method. The data processing method comprises determining a structure type of the upload data according to the data content; determining a first table of a first database according to the structure type, the data content and the table name of the upload data; and storing the upload data in the first table of the first database; wherein the first database is the relational database or the NoSQL database.

Abstract: In an embodiment, a method includes forming a plurality of semiconductor fins over a substrate, the plurality of semiconductor fins comprising a first fin, a second fin, a third fin, and a fourth fin; forming a first dielectric layer over the plurality of semiconductor fins, the first dielectric layer filling an entirety of a first trench between the first fin and the second fin; forming a second dielectric layer over the first dielectric layer, the second dielectric layer filling an entirety of a second trench between the second fin and the third fin, the forming the second dielectric layer comprising: forming an oxynitride layer; and forming an oxide layer; and forming a third dielectric layer over the second dielectric layer, the third dielectric layer filling an entirety of a third trench between the third fin and the fourth fin.

Abstract: A data processing method applied in a data center is provided.

Abstract: A data searching method for a data dictionary, comprising obtaining an input word and performing a word embedding operation on the input word to generate a word vector of the input word, calculating a degree of correlation between the input word and a plurality of words in the data dictionary according to the word vector of the input word, and determining at least one recommendation synonym from the plurality of words in the data dictionary according to the degree of correlation between the input word and the plurality of words in the data dictionary, wherein the step including when determining that the cosine similarity between the word vector of the input word and a word vector of a first word of the plurality of words in the data dictionary is greater than a threshold value, determining the first word as a recommendation synonym.

Abstract: A sense amplifier of a memory device that includes sense amplifier circuits and a reference sharing circuit is introduced. The sense amplifier circuits are configured to sense the plurality of bit lines according to an enable signal. The reference sharing circuit includes first switches and second switches that are coupled to the reference nodes and second reference nodes of the sense amplifier circuits, respectively. The first switches and second switches are controlled according to a control signal to control a first electrical connection among the first reference nodes, and to control a second electrical connection among the second reference nodes. An operation method of the sense amplifier and a memory device including the sense amplifier are also introduced.

Abstract: The sense amplifier circuit includes a differential amplifier, a first switch, and a second switch. The differential amplifier includes a first input node, a second input node, a first output node, and a second output node. The differential amplifier amplifies a voltage difference of the first output node and the second output node according to a first input voltage of the first input node and a second input voltage of the second input node. A control node of the first (second) switch is coupled to a control line, the first (second) switch is coupled to the first (second) input node, and the first (second) switch is coupled to the first (second) output node. The first (second) switch pre-charges the first (second) input node by a first (second) output voltage of the first (second) output node while the control line is received a select signal.

Abstract: The invention discloses a pure end-to-end deep reinforcement learning for training car racing game AI bot that uses only the velocity information extracted from screen for both training and testing phases without using any internal state from game environment, such as the car facing angle. The learned AI bot can play better than the average performance of human players. In addition, the reward function is designed to consist only the velocity value, and use Ape-X distributed training framework combined with a variant of Deep Q Network to solve the sparse training signal problem caused by the reward function of an original design. Moreover, limit learner rate method is designed that improves the training efficiency and training performance. The AI bot trained in this way can achieve performance beyond the average human level and reach a level close to professional players.

Abstract: Provided are devices and methods for forming devices. An exemplary method includes etching a cavity in a vertical direction into a fin structure including at least one semiconductor nanosheet overlying a sacrificial layer, wherein the cavity is formed with a sidewall; recessing the sacrificial layer by a lateral distance to a recessed surface; forming an inner spacer laterally adjacent to the recessed surface of the sacrificial layer, wherein the inner spacer has a lateral width greater than the lateral distance; and growing epitaxial material in the cavity to form a source/drain region laterally adjacent to the inner spacer.

Abstract: Embodiment described herein provide systems, apparatuses and methods for convoluting a filter (“kernel”) to input data in the form of an input array by reusing computations of repeated data entries in the input array due to convolution movements from one convolution step to the next. In one embodiment, to compute a convolution of an input matrix and a filter matrix, instead of unrolling data entries from the input matrix of each convolution step into an input vector, only non-repeated new data entries at each convolution step may be added to the input vector. An input mapping circuit that implements an input parameter mapping matrix may then iteratively map data entries of the input vector to different weight registers that corresponds to weights in the filter matrix.

Abstract: An immersion cooling system includes an immersion tank and at least one electronic device. The immersion tank is configured to accommodate a coolant. The at least one electronic device is disposed in the immersion tank, and includes a cage assembly, at least one motherboard and a plurality of expansion cards. The cage assembly includes a fixed frame and at least one movable frame. The at least one movable frame are movably disposed on the fixed frame. The at least one motherboard is mounted on the at least one movable frame. The plurality of expansion cards are mounted on the at least one motherboard.

Abstract: A coil module is provided, including a second coil mechanism. The second coil mechanism includes a third coil assembly and a second base corresponding to the third coil assembly. The second base has a positioning assembly corresponding to a first coil mechanism.

Abstract: A plastic light-folding element includes an incident surface, an exit surface, at least one reflective surface, at least one connecting surface and at least one gate vestige structure. The incident surface is configured to lead an imaging light enter the plastic light-folding element. The exit surface is configured to lead the imaging light exit the plastic light-folding element. The reflective surface is configured to fold the imaging light. The connecting surface is connected to the incident surface, the exit surface and the reflective surface. The gate vestige structure is disposed on the connecting surface. At least one of the incident surface, the exit surface and the reflective surface includes an optical portion and an arc step structure, the arc step structure is disposed on a periphery of the optical portion, and an arc is formed by the arc step structure centered on the optical portion.

Abstract: A driving mechanism is provided for moving an optical element, including a fixed module, a movable module holding the optical element, a driving assembly for driving the movable module to move relative to the fixed module, a position-sensing element, and a 3D circuit. The fixed module has a base, and the position-sensing element is disposed on the base to detect the movement of the movable module relative to the fixed module. The 3D circuit is embedded in the base and electrically connected to the position-sensing element.

Abstract: A memory structure including a substrate, charge storage layers, and a gate is provided. The charge storage layers are located on the substrate. The gate is located on the substrate on one side of the charge storage layers. The gate extends along a first direction. The gate has a protruding portion protruding along a second direction. The second direction intersects the first direction. The protruding portion is located between two adjacent charge storage layers arranged along the first direction.

Abstract: A display apparatus includes a circuit substrate, light-emitting elements, a bonding pad and an encapsulation glue layer. The circuit substrate has a display area, an outer lead bonding area and a connection area connected between the display area and the outer lead bonding area. The light emitting elements are disposed on the display area of the circuit substrate. The bonding pad is disposed on the outer lead bonding area of the circuit substrate. The encapsulation glue layer is disposed on the circuit substrate and includes a flat portion and a thick wall portion connected with each other. The flat portion covers the light emitting elements. The thick wall portion extends to the connection area. The thick wall portion has a thickness which gradually increases toward the outer lead bonding area. In addition, an array substrate is also provided.

Abstract: An optical element driving mechanism is provided, including a fixed part, a movable part, a metallic member and a driving assembly. The fixed part includes a base. The movable part is movably connected to the fixed part, and carries an optical element, the optical element has an optical axis. The metallic member is disposed on the base, and includes an inner electrical connection part and an outer electrical connection part, the inner electrical connection part and the outer electrical connection part are connected to each other. The driving assembly includes at least one driving magnetic element and drives the movable part to move relative to the fixed part.

Abstract: Semiconductor devices and methods of forming the same are provided. In one embodiment, a semiconductor device includes an active region including a channel region and a source/drain region and extending along a first direction, and a source/drain contact structure over the source/drain region. The source/drain contact structure includes a base portion extending lengthwise along a second direction perpendicular to the first direction, and a via portion over the base portion. The via portion tapers away from the base portion.

Abstract: A lamination apparatus is adapted to laminate an encapsulation glue and a panel. The lamination apparatus includes a lower mold, an upper mold, and a cutting knife. The lower mold has a lamination area and a peripheral area adjacent to the lamination area. The panel is adapted to be retained in the lamination area. The upper mold is movable close to or away from the lower mold, and includes an encapsulation glue fixing part. The encapsulation glue fixing part is adapted to fix the encapsulation glue to the upper mold. The cutting knife is disposed in the peripheral area of the lower mold, and is located in a horizontal direction between the lamination area of the lower mold and the encapsulation glue fixing part of the upper mold. In addition, a lamination method is also disclosed.

Abstract: A greenhouse gas verification system includes a local data verification device, an upstream data verification device, and a data weighting device. The local data verification device obtains multiple sensor data and multiple data coefficients associated with the sensor data, and calculates a total greenhouse gas quantity according to these sensor data and data coefficients. The upstream data verification device obtains a greenhouse gas value and upstream evidence associated with the greenhouse gas value, and generates a verification result according to the upstream evidence and the greenhouse gas value. When the verification result is passed, the upstream data verification device outputs the greenhouse gas values. The data weighting device receives the total greenhouse gas quantity and the greenhouse gas value, and performs weighted calculations according to the greenhouse gas value and total greenhouse gas quantity to output a greenhouse gas value vector.

Abstract: A power consumption monitoring device includes a sensor, a storage, and a processor. The sensor is configured to detect a power-consuming device quantity and a power consumption amount. The storage is configured to store the power-consuming device quantity and the power consumption amount. The processor is communicatively connected to the sensor and the storage. The processor is configured to calculate a power-consuming device idling indicator based on the power-consuming device quantity and the power consumption amount in a monitoring time interval, wherein the power-consuming device idling indicator is used for indicating a deviation status of the power-consuming device quantity and the power consumption amount. The processor is further configured to determine whether the power-consuming device idling indicator exceeds a warning threshold. In response to the power-consuming device idling indicator exceeding the warning threshold, the processor is further configured to generate a warning message.