Abstract: Vision Transformers (ViT) have shown their competitive advantages performance-wise compared to convolutional neural networks (CNNs) though they often come with high computational costs. Methods, systems and techniques herein learn instance-dependent attention patterns, utilizing a lightweight connectivity predictor module to estimate a connectivity score of each pair of tokens. Intuitively, two tokens have high connectivity scores if the features are considered relevant either spatially or semantically. As each token only attends to a small number of other tokens, the binarized connectivity masks are often very sparse by nature providing an opportunity to accelerate the network via sparse computations. Equipped with the learned unstructured attention pattern, sparse attention ViT produces a superior Pareto-optimal trade-off between FLOPs and top-1 accuracy on ImageNet compared to token sparsity (48%˜69% FLOPs reduction of MHSA; accuracy drop within 0.4%).

Abstract: A composite phosphor ceramic optical fiber with luminous efficiency and color rendering index includes a phosphor ceramic fiber core and a phosphor ceramic cladding, the phosphor ceramic fiber core uses a yellow phosphor ceramic doped with Ce ions, and the phosphor ceramic cladding uses a red phosphor ceramic co-doped with ions. The composite phosphor ceramic fiber is prepared by the gel casting. With this design structure, not only can the energy transfer between Ce ions and other ions cause a red shift in the spectrum to improve the color rendering index, but also the luminous efficiency can be significantly enhanced. In addition, a diameter of the composite phosphor ceramic optical fiber matches a laser spot size very well, effectively solving the “yellow ring effect” problem caused by the mismatch. It can effectively solve the heat dissipation problem during high power LD pumping and is conducive to long term stable lighting.

Abstract: Vision Transformers (ViT) have shown their competitive advantages performance-wise compared to convolutional neural networks (CNNs) though they often come with high computational costs. Methods, systems and techniques herein learn instance-dependent attention patterns, utilizing a lightweight connectivity predictor module to estimate a connectivity score of each pair of tokens. Intuitively, two tokens have high connectivity scores if the features are considered relevant either spatially or semantically. As each token only attends to a small number of other tokens, the binarized connectivity masks are often very sparse by nature providing an opportunity to accelerate the network via sparse computations. Equipped with the learned unstructured attention pattern, sparse attention ViT produces a superior Pareto-optimal trade-off between FLOPs and top-1 accuracy on ImageNet compared to token sparsity (48%˜69% FLOPs reduction of MHSA; accuracy drop within 0.4%).

Abstract: The invention relates to a method for determining an integrated network loss rate in the UHV AC cross-regional electricity trading. The method includes fitting a curve relationship between integrated network loss rates and transmitting powers on the basis of theoretical calculations of the UHV AC transmission line loss, using a relational fitted curve between actual values of integrated network loss rates and transmitting powers calculated according to gateway power statistical data to perform geometrical average correction on the original curve, and making planned values of the integrated network loss rates to be closer to the actual values according to a method for determining UHV AC integrated network loss rates according to a correction curve function relation, which greatly increases fairness of the trading settlement. The method is simple and easy to implement with high accuracy, and applicable to planning and trading settlement of regular or real-time UHV AC electricity trading.

Abstract: A gamma-ray analysis system is described for analyzing gamma-ray emitting radionuclides. The gamma-ray analysis system includes an analytical apparatus having a gamma-ray detector in operative communication with a modular and scalable shield assembly that encases a sample container having a sample to be tested. The detector communicates data to an electronic interface device that converts the data from an analog format to a digital format before a controller transmits the data to a central laboratory for further data processing, analysis and conclusion by qualified laboratory analysts. The controller runs an application software package on a graphic user interface that allows simple steps for conducting testing and data acquisition by the end user, while permitting real time data transmission between the field site and the central location. Functions were implemented for ensuring laboratory quality results while removing knowledge and experience requirements of an end user.

Abstract: A gamma-ray analysis system is described for analyzing gamma-ray emitting radionuclides. The gamma-ray analysis system includes an analytical apparatus having a gamma-ray detector in operative communication with a modular and scalable shield assembly that encases a sample container having a sample to be tested. The detector communicates data to an electronic interface device that converts the data from an analog format to a digital format before a controller transmits the data to a central laboratory for further data processing, analysis and conclusion by qualified laboratory analysts. The controller runs an application software package on a graphic user interface that allows simple steps for conducting testing and data acquisition by the end user, while permitting real time data transmission between the field site and the central location. Functions were implemented for ensuring laboratory quality results while removing knowledge and experience requirements of an end user.

Abstract: A method including receiving to-be-queried domain names; obtaining a query capacity of a preset query group; allocating the domain names to a matching query group according to the query capacity; and querying domain name information of the domain names through a query interface of the matching query group. In one aspect, by flexible domain name allocation, query resources are fully utilized, and the quantity of concurrent domain name queries is increased, thereby improving query efficiency and reducing a query time. In another aspect, queries are performed by using a particular query interface, and settings of the query interface are maintained, thereby avoiding that settings of the query interface need to be switched due to queries for different types of domain names, and improving query efficiency and reducing a query time.

Abstract: The invention relates to a method for determining an integrated network loss rate in the UHV AC cross-regional electricity trading. The method includes fitting a curve relationship between integrated network loss rates and transmitting powers on the basis of theoretical calculations of the UHV AC transmission line loss, using a relational fitted curve between actual values of integrated network loss rates and transmitting powers calculated according to gateway power statistical data to perform geometrical average correction on the original curve, and making planned values of the integrated network loss rates to be closer to the actual values according to a method for determining UHV AC integrated network loss rates according to a correction curve function relation, which greatly increases fairness of the trading settlement. The method is simple and easy to implement with high accuracy, and applicable to planning and trading settlement of regular or real-time UHV AC electricity trading.

Abstract: An apparatus for measuring ammonia gas concentration in an ongoing chemical mechanical polishing (CMP) cycle utilizing an acidic CMP slurry, having the following components: a. A transferring means to collect a sample of the acidic CMP slurry; b. A converting means to convert the acidic CMP slurry to a basic slurry; c. A measuring means to measure the ammonia gas present in the basic slurry; d. A detection means to signal the end of an ongoing CMP cycle.

Abstract: An apparatus for measuring ammonia gas concentration in an ongoing mechanical polishing (CMP) cycle utilizing an acidic CMP slurry, having the following components: a. A transferring means to collect a sample of the acidic CMP slurry; b. A converting means to convert the acidic CMP slurry to a basic slurry; c. A measuring means to measure the ammonia gas present in the basic slurry; d. A detection means to signal the end of an ongoing CMP cycle.

Abstract: An instrumentation device for controlling one or more instruments, wherein the instrumentation device having an interface which accepts at least three sets of inputs and transmits at least three sets of outputs, the interface capable of transmitting signals of different voltage between the individual inputs and individual outputs of the interface and wherein the interface is capable of accepting, translating and transmitting as one of the at least three sets of outputs, input from more than one set of the at least sets of three inputs.

Abstract: An apparatus is described for detecting the presence of a gaseous chemical produced during a chemical-mechanical polishing operation. The apparatus includes a catalytic converter, a reaction chamber and a light sensor. The catalytic converter, heated to about 800° C. converts the chemical to a different chemical product. The reaction chamber produces an excited species; the pressure in the reaction chamber is maintained sufficiently low to substantially avoid collisional deactivation of the excited species, so as to permit real-time detection of the chemical. A light signal from the excited species is input to the light sensor. An output from the light sensor corresponds to the real-time detection of the chemical, thereby permitting real-time control of the chemical-mechanical polishing operation.

Abstract: Detection of the endpoint for removal of a target film overlying a stopping film by removing the target film with a process that selectively generates a chemical reaction product (for example ammonia when polishing a wafer with a nitride film in a slurry containing KOH) with one of the stopping film and the target film, converting the chemical reaction product to a separate product, producing excited molecules from the separate product, and monitoring the level of light emitted from the excited molecules as the target film is removed.

Abstract: Detection of the endpoint for removal of a target film overlying a stopping film by removing the target film with a process that selectively generates a chemical reaction product (for example, ammonia when polishing a wafer with a nitride film in a slurry containing KOH) with either the target or stopping film, and monitoring the level of chemical reaction product as the target film is removed. The reaction product is extracted as a gas from the slurry and monitored using a threshold photoionization mass spectrometer.

Abstract: A method is described for fabricating a cloisonné structure, in which a top surface of a metal oxide layer is made coplanar with a top surface of a metallic structure formed on a substrate. A nitride layer is deposited on at least the top surface of the metallic structure, and the metal oxide layer is deposited over the metallic structure and the nitride layer. The metal oxide layer is then polished by a chemical-mechanical polishing (CMP) process using a slurry, to expose the nitride layer on the top surface of the metallic structure. Polishing of the nitride layer causes ammonia to be generated in the slurry. The ammonia is extracted as a gas from the slurry, and a signal is generated in accordance with the ammonia concentration. The CMP process is terminated in accordance with a change in the signal. In a preferred embodiment, the metal oxide is aluminum oxide, the nitride is aluminum nitride, and the nitride layer is deposited as a conformal layer on the substrate and the metallic structure.

Abstract: Determination of an endpoint for removing a film from a wafer, by determining a first reference point removal time indicating when a breakthrough of the film has occurred, determining a second reference point removal time indicating when the film has been polished almost to completion, determining an additional removal time indicating an overpolishing interval, and adding the second reference point removal time with the additional removal time to get a total removal time to the endpoint.

Abstract: In a chemical mechanical planarization assembly directed for the removal of oxide layers, which stop on films containing silicon nitride, a conventional polishing table is provided with a gas extraction unit which transmits a gas sample to an infrared spectrometer. The presence of ammonia in the slurry, which is generated when a stop layer containing silicon nitride is abraded under high pH conditions, can be detected using infrared spectroscopy and accordingly provides for an in situ endpoint detection method.

Abstract: The invention discloses an apparatus for detecting the presence of a chemical in a gaseous state, having:

Abstract: A method is provided to optimize the chemical mechanical planarization process of wafers having a target removal layer and a stop layer. A slurry is added to a polishing table which includes a polishing pad and a platen adapted for rotation; a portion of the slurry is allowed to engage an interface between the polishing pad and the wafer. A gaseous sample is continuously extracted from the slurry; the gaseous sample includes a reactant product created when the polishing pad engages the stop layer. The gaseous sample is introduced into a reactant product detector. A first time is determined, corresponding to an initial detection of the reactant product in the slurry, thereby creating a first reference point. A second time is determined, corresponding to the detection of a maximum volume of the reactant in said slurry, thereby creating a second reference point.

Abstract: A method and apparatus are described for detecting an endpoint of a film removal process in which a target film overlying a stopping film is removed. A chemical reaction product is generated from at least one of the target film and the stopping film; this chemical reaction product is converted to a separate product. The separate product is exposed to ionizing radiation. The ionization current generated by the radiation is monitored as the target film is removed. A change in the ionization current corresponds to a change in concentration of the separate product, thereby indicating the endpoint of the film removal process. In the particular case of removal of a silicon dioxide film overlying a silicon nitride film by chemical-mechanical polishing, the reaction product is ammonia extracted from the polishing slurry. The ammonia is converted to ammonium chloride by a reaction with hydrogen chloride vapor.