Abstract: A manufacturing method of a thin-film power inductor includes: Alloy powder is mixed with plasticizer, adhesive, curing agent, dispersing agent and organic solvent to form slurry; the slurry is applied on a PET film, and drying to form a magnetic band; and the magnetic band is cut to form a plurality of magnetic sheets. A hole is opened on a magnetic sheet to form a hole-shaped magnetic sheet. Electrodes are processed on an insulating substrate to form a coil layer. Magnetic sheets, hole-shaped magnetic sheets, and the coil layer are stacked to form a block. The block is pressed, and the block is cut to form an individual product. The individual product is baked to form a main body. Silver paste is applied on the main body to form outer electrodes. A nickel layer and a tin layer are electroplated on outer electrodes to form a thin-film power inductor.

Abstract: A catalyst contains a metal oxide, and a molecular sieve, in a crystal form, having a topological pore structure. The metal oxide is centrally distributed on the surface of the molecular sieve. Grains of the molecular sieve are exposed to at least three families of crystal planes. The family of crystal plane with the largest pore size in topology is occupied by the metal oxide by no more than 30%, preferably no more than 20%, or no more than 10%.

Abstract: An integrated catalyst can be used in an olefin disproportionation reaction. The integrated catalyst contains a plurality of different integrated active phases. The relative positions among different active phases remain substantially unchanged during the olefin disproportionation reaction. The effective distance between respective bisecting planes of two adjacent different active phases is 0.5-5 mm, preferably 1-3 mm.

Abstract: The invention relates to a molecular sieve composition, a process of preparing same and use thereof in the production of lower olefins. The molecular sieve composition comprises an aluminophosphate molecular sieve and a CO adsorbing component, both of which are present independently of each other. When the molecular sieve composition is used as a catalyst for producing lower olefins using synthesis gas as a raw material, the molecular sieve composition has the advantages of high selectivity to lower olefins and the like.

Abstract: A process for producing ethylene includes the steps of: (1) catalytically cracking a raw material hydrocarbon to obtain a first stream containing propylene; (2) separating the first stream to obtain a C3 component stream; and (3) disproportionating the C3 component stream to obtain an ethylene stream. This process reduces the overall energy consumption of the production process and increases the total yield of the ethylene while obtaining a polymerization-grade ethylene.

Abstract: A process for producing ethylene includes the steps of: (1) catalytically cracking a raw material hydrocarbon to obtain a first stream containing propylene; (2) separating the first stream to obtain a C3 component stream; and (3) disproportionating the C3 component stream to obtain an ethylene stream. This process reduces the overall energy consumption of the production process and increases the total yield of the ethylene while obtaining a polymerization-grade ethylene.

Abstract: The invention relates to a molecular sieve composition, a process of preparing same and use thereof in the production of lower olefins. The molecular sieve composition comprises an aluminophosphate molecular sieve and a CO adsorbing component, both of which are present independently of each other. When the molecular sieve composition is used as a catalyst for producing lower olefins using synthesis gas as a raw material, the molecular sieve composition has the advantages of high selectivity to lower olefins and the like.

Abstract: A method includes storing parameter versions utilized by learner instances in each of two or more epochs in a parameter receiving buffer of a parameter server, the learner instances performing distributed execution of workload computations of a machine learning algorithm. The method also includes creating a parameter roster in the parameter server comprising parameter version vectors specifying the parameter versions used by each of the learner instances during each of the two or more epochs. The method further includes generating one or more aggregated parameter sets for storage in an aggregated parameters buffer by aggregating parameter versions from the parameter receiving buffer based on the parameter version vectors in the parameter roster and providing aggregated parameter sets from the aggregated parameters buffer to the learner instances for deterministic replay of the distributed execution of the workload computations of the machine learning algorithm.

Abstract: A set of instructions stored on at least one computer readable medium for running on a computer system. The set of instructions includes instructions for identifying line segments of a roof that is preferably displayed within a geo-referenced image, instructions for determining position, orientation and length of the line segments utilizing the geo-referenced image, and instructions for classifying, automatically, at least one of the line segments as one of a plurality of predefined roof elements utilizing at least one of the relative position and orientation of the line segments.