Abstract: A centrifugal heat dissipation fan including a housing and an impeller is provided. The housing has at least one inlet disposed along an axis and at least one first outlet and a second outlet located in different radial directions, wherein the first outlet and the second outlet are opposite to and separated from each other. The impeller is disposed in the housing along the axis. A heat dissipation system of an electronic device is also provided.

Abstract: An automatic nucleic acid detection system and a method thereof are disclosed. The automatic nucleic acid detection method includes: performing, by an automatic control subsystem, on a nucleic acid extraction machine platform, a nucleic acid extraction on one or more specimens in a sample tray to generate one or more corresponding nucleic acids in the sample tray; distributing, by the automatic control subsystem, on a nucleic acid distribution machine platform, the nucleic acid in each hole of the sample tray and a first reagent into a plurality of holes of a detection tray, wherein the number of holes of the detection tray is greater than that of the sample tray; and performing, by the automatic control subsystem, on a nucleic acid detection machine platform, a nucleic acid detection on the detection tray.

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 phantom-free calibration method for a computerized tomography scan has the steps of: scanning a plurality of homogeneous human tissues with a computerized tomography scanner and obtaining homogeneous human tissue scan information; scanning a tissue to be tested with the computerized tomography scanner and obtaining tissue to be tested scan information; calculating a spectral characteristic parameter of the computerized tomography scanner with a computing device using a model, wherein a standard tissue parameter of the homogeneous human tissues and the human tissue scan information are used in the model; and calculating a tissue parameter of the tissue to be tested with the computing device using the model, wherein the spectral characteristic parameter and the tissue to be tested scan information are inputted in the model. The phantom-free calibration method makes the tissue parameters of the tissue to be tested more accurate and the calibration process easier and more convenient.

Abstract: A sense amplifier circuit includes a sense amplifier, a switch and a temperature compensation circuit. The temperature compensation circuit provides a control signal having a positive temperature coefficient, based on which the switch provides reference impedance for temperature compensation. The sense amplifier includes a first input end coupled to a target bit and a second input end coupled to the switch. The sense amplifier outputs a sense amplifier signal based on the reference impedance and the impedance of the target bit.

Abstract: Semiconductor device and the manufacturing method thereof are disclosed herein. An exemplary semiconductor device comprises a substrate; semiconductor layers over the substrate, wherein the semiconductor layers are separate from each other and are stacked up along a direction generally perpendicular to a top surface of the substrate; a dielectric feature over and separate from the semiconductor layers; and a gate structure wrapping around each of the semiconductor layers, the gate structure having a gate dielectric layer and a gate electrode layer, wherein the gate dielectric layer interposes between the gate electrode layer and the dielectric feature and the dielectric feature is disposed over at least a part of the gate electrode layer.

Abstract: A fan is provided herein, including a housing, a hub, and a plurality of blades. The housing includes a top case and a bottom case. The hub is rotatably disposed between the top case and the bottom case in an axial direction. The blades extend from the hub in a radial direction, located between the top case and the bottom case. Each of the blades has a proximal end and a distal end. The proximal end is connected to the hub. The distal end is opposite from the proximal end, located at the other side of the blade, having at least one recessed portion. Each of the recessed portions form a passage for air.

Abstract: The present disclosure provides intelligent radio frequency interference mitigation in a computing platform. The computing platform includes a processor, a memory, a system clock and a wireless network interface. The system clock can be controlled so that the processor and/or the memory may operate at a slow frequency or a fast frequency. The wireless network may operate on a radio channel that experiences radio frequency interference at the fast frequency. The system clock may be intelligently controlled to select the slow frequency to reduce radio frequency interference to prioritize execution of a network application, or to select the fast frequency to increase processor speed and prioritize execution of a local application.

Abstract: A bottom-pinned spin-orbit torque magnetic random access memory (SOT-MRAM) is provided in the present invention, including a substrate, a bottom electrode layer on the substrate, a magnetic tunnel junction (MTJ) on the bottom electrode layer, a spin-orbit torque (SOT) layer on the MTJ, a capping layer on the SOT layer, and an injection layer on the capping layer, wherein the injection layer is divided into individual first part and second part, and the first part and the second part are connected respectively with two ends of the capping layer.

Abstract: A computing system performs shared cache allocation to allocate cache resources to groups of tasks. The computing system monitors the bandwidth at a memory hierarchy device that is at a next level to the cache in a memory hierarchy of the computing system. The computing system estimates a change in dynamic power from a corresponding change in the bandwidth before and after the cache resources are allocated. The allocation of the cache resources are adjusted according to an allocation policy that receives inputs including the estimated change in the dynamic power and a performance indication of task execution.

Abstract: A computing system performs partial cache deactivation. The computing system estimates the leakage power of a cache based on operating conditions of the cache including voltage and temperature. The computing system further identifies a region of the cache as a candidate for deactivation based on cache hit counts. The computing system then adjusts the size of the region for the deactivation based on the leakage power and a bandwidth of a memory hierarchy device. The memory hierarchy device is at the next level to the cache in a memory hierarchy of the computing system.

Abstract: A method for evaluating a risk of a subject getting a specific disease includes steps of: storing a reference database that contains original parameter sets; selecting target alleles from an SNP profile derived from genome sequencing data of a subject; selecting target parameter sets from among the original parameter sets; calculating, for each of the target parameter sets, a race factor based on a global risk allele frequency and a group-specific risk allele frequency included in the target parameter set; calculating a genetic factor based on statistics, global reference allele frequencies, the race factors for the target parameter sets, and numbers of chromosomes in homologous chromosome pairs included in the target parameter sets; calculating a citation factor based on numbers of citation times included in the target parameter sets; and calculating a risk score based on the genetic factor and the citation factor.

Abstract: Provided is an anti-fuse memory including a anti-fuse memory cell including an isolation structure, a select gate, first and second gate insulating layers, an anti-fuse gate, and first, second and third doped regions. The isolation structure is disposed in a substrate. The select gate is disposed on the substrate. The first gate insulating layer is disposed between the select gate and the substrate. The anti-fuse gate is disposed on the substrate and partially overlapped with the isolation structure. The second gate insulating layer is disposed between the anti-fuse gate and the substrate. The first doped region and the second doped region are disposed in the substrate at opposite sides of the select gate, respectively, wherein the first doped region is located between the select gate and the anti-fuse gate. The third doped region is disposed in the substrate and located between the first doped region and the isolation structure.

Abstract: Aspects of the present disclosure provide a method for training a predictor that predicts performance of a plurality of machine learning (ML) models on platforms. For example, the method can include converting each of the ML models into a plurality of instructions or the instructions and a plurality of intermediate representations (IRs). The method can also include simulating execution of the instructions corresponding to each of the ML models on a platform and generating instruction performance reports. Each of the instruction performance reports can be associated with performance of the instructions corresponding to one of the ML models that are executed on the platform. The method can also include training the predictor with the instructions or the IRs as learning features and the instruction performance reports as learning labels, compiling the predictor into a library file, and storing the library file in a storage device.

Abstract: According to an exemplary embodiment, a method of forming a vertical device is provided. The method includes: providing a protrusion over a substrate; forming an etch stop layer over the protrusion; laterally etching a sidewall of the etch stop layer; forming an insulating layer over the etch stop layer; forming a film layer over the insulating layer and the etch stop layer; performing chemical mechanical polishing on the film layer and exposing the etch stop layer; etching a portion of the etch stop layer to expose a top surface of the protrusion; forming an oxide layer over the protrusion and the film layer; and performing chemical mechanical polishing on the oxide layer and exposing the film layer.

Abstract: A centrifugal heat dissipation fan including a housing and an impeller disposed in the housing on an axis is provided. The housing has at least one inlet on the axis and has a plurality of outlets in different radial directions. A heat dissipation system of an electronic device is also provided.

Abstract: The present disclosure describes an apparatus with a local interconnect structure. The apparatus can include a first transistor, a second transistor, a first interconnect structure, a second interconnect structure, and a third interconnect structure. The local interconnect structure can be coupled to gate terminals of the first and second transistors and routed at a same interconnect level as reference metal lines coupled to ground and a power supply voltage. The first interconnect structure can be coupled to a source/drain terminal of the first transistor and routed above the local interconnect structure. The second interconnect structure can be coupled to a source/drain terminal of the second transistor and routed above the local interconnect structure. The third interconnect structure can be routed above the local interconnect structure and at a same interconnect level as the first and second interconnect structures.

Abstract: A semiconductor structure includes a die embedded in a molding material, the die having die connectors on a first side; a first redistribution structure at the first side of the die, the first redistribution structure being electrically coupled to the die through the die connectors; a second redistribution structure at a second side of the die opposing the first side; and a thermally conductive material in the second redistribution structure, the die being interposed between the thermally conductive material and the first redistribution structure, the thermally conductive material extending through the second redistribution structure, and the thermally conductive material being electrically isolated.

Abstract: In an embodiment, a package structure including an electro-optical circuit board, a fanout package disposed over the electro-optical circuit board is provided. The electro-optical circuit board includes an optical waveguide. The fanout package includes a first optical input/output portion, a second optical input/output portion and a plurality of electrical input/output terminals electrically connected to the electro-optical circuit board. The first optical input/output portion is optically coupled to the second optical input/output portion through the optical waveguide of the electro-optical circuit board.

Abstract: In an embodiment, a package structure including an electro-optical circuit board, a fanout package disposed over the electro-optical circuit board is provided. The electro-optical circuit board includes an optical waveguide. The fanout package includes a first optical input/output portion, a second optical input/output portion and a plurality of electrical input/output terminals electrically connected to the electro-optical circuit board. The first optical input/output portion is optically coupled to the second optical input/output portion through the optical waveguide of the electro-optical circuit board.