Abstract: Embodiments of this specification provide a method and an apparatus for deploying a federated learning task based on a container. In the method, a federated learning task is deployed to a plurality of service party devices by using a container management platform, and the federated learning task is executed by using the plurality of service party devices. In the method, when receiving a task description file for the federated learning task, the container management platform can respectively generate first container group description files for the plurality of service party devices based on the task description file, and respectively send the plurality of generated first container group description files to the corresponding service party devices. The plurality of service party devices create container groups based on the respectively received first container group description files, and execute the federated learning task by using the created container groups.

Abstract: Disclosed in the present invention are TRIP steel and a preparation method therefor, a cold-rolled steel sheet, and a hot-dip galvanized steel sheet, wherein the TRIP steel comprises the following chemical components, in percentages by mass: C: 0.15-0.3%, Si: 0.6-1.0%, Mn: 1.7-2.5%, Al: 0.5-0.9%, P?0.01%, S?0.01%, N?0.007%, and Fe?90%. The TRIP steel of the present invention is based on carbon-silicon-manganese steel. By means of optimizing the proportions of carbon, silicon, manganese and aluminum, and replacing Si in traditional TRIP steel with a part of the Al, the effects of inhibiting carbide precipitation and stabilizing residual austenite can be achieved so as to ensure the mechanical properties of the TRIP steel, and the surface quality and plateability of the steel sheet can also be improved.

Abstract: A device includes a substrate, a gate structure wrapping around a vertical stack of nanostructure semiconductor channels, and a source/drain abutting the vertical stack and in contact with the nanostructure semiconductor channels. The device includes a gate via in contact with the first gate structure. The gate via includes a metal liner layer having a first flowability, and a metal fill layer having a second flowability higher than the first flowability.

Abstract: A charging control circuit is provided. The charging control circuit includes a battery module, an oscillator, a charge pump, and a first switch circuit. The battery module is configured to provide a power supply voltage in a shipping mode. The oscillator is configured to generate an oscillation signal according to the power supply voltage. The charge pump is configured to receive the power supply voltage and boost the power supply voltage in response to the oscillation signal, thereby providing an output voltage to a positive terminal of the battery module. The first switch circuit is coupled between the battery module, the oscillator, and the charge pump and configured to be turned on according to a first switch signal to provide the power supply voltage to the oscillator and the charge pump.

Abstract: Embodiments of this specification provide a compilation method and a compiler for compiling a script of a privacy-preserving machine learning algorithm. In the compilation method, a compiler obtains a description script written in a predetermined format. The description script defines at least a computing formula in a privacy-preserving machine learning algorithm. The compiler determines several privacy algorithms for executing several operators used in the computing formula; obtains several code modules for executing the several privacy algorithms; and generates program code corresponding to the description script based on the several code modules.

Abstract: In some embodiments, the present disclosure relates to an integrated circuit device. A transistor structure includes a gate electrode separated from a substrate by a gate dielectric and a pair of source/drain regions disposed within the substrate on opposite sides of the gate electrode. A lower conductive plug is disposed through a lower inter-layer dielectric (ILD) layer and contacting a first source/drain region. A capping layer is disposed directly on the lower conductive plug. An upper inter-layer dielectric (ILD) layer is disposed over the capping layer and the lower ILD layer. An upper conductive plug is disposed through the upper ILD layer and directly on the capping layer.

Abstract: The present disclosure provides an active metal-containing M-CHA/M-MOR composite molecular sieve and a preparation method. The preparation method includes: S1, adding an alkali source, a silicon source, an aluminum source, a structure-directing agent, an active metal complex and a crystal form modifier into water, and performing stirring to obtain a gel, wherein the crystal form modifier is an aqueous solution containing a molecular sieve secondary structural unit prepared based on at least one of an MOR molecular sieve, a CHA/MOR composite molecular sieve, and a CHA-MOR mixed molecular sieve; S2, subjecting the gel to a hydrothermal reaction, and separating the resulting slurry to obtain a composite molecular sieve; and S3, drying the composite molecular sieve, and performing calcining to obtain the active metal-containing M-CHA/M-MOR composite molecular sieve. The direct preparation method has the advantages of short process flow, low wastewater production, low energy consumption and the like.

Abstract: The present disclosure describes a method for the fabrication of ruthenium conductive structures over cobalt conductive structures. In some embodiments, the method includes forming a first opening in a dielectric layer to expose a first cobalt contact and filling the first opening with ruthenium metal to form a ruthenium contact on the first cobalt contact. The method also includes forming a second opening in the dielectric layer to expose a second cobalt contact and a gate structure and filling the second opening with tungsten to form a tungsten contact on the second cobalt contact and the gate structure. Further, the method includes forming a copper conductive structure on the ruthenium contact and the tungsten contact, where the copper from the copper conductive structure is in contact with the ruthenium metal from the ruthenium contact.

Abstract: The present invention is directed to a peptide, multimer, conjugate, analog, derivative or mimetic thereof that inhibits the activity of VISTA. The invention further contemplates therapeutic use of the VISTA antagonist peptide, multimer, conjugate, derivative or mimetic thereof, including treating or preventing cancer, bacterial infections, viral infections, parasitic infections and fungal infections, as well as research uses of the antagonist.

Abstract: A fan, comprising blades (102), an upper hub (111), and a lower hub (112). Each blade (102) comprises a connecting part (211). The connecting part (211) is clamped between a lower surface of the upper hub (111) and an upper surface of the lower hub (112). The lower surface of the upper hub (111) is provided with an upper hub connecting area, and the upper surface of the lower hub (112) is provided with a lower hub connecting area. An upper surface and a lower surface of the connecting part (211) are separately provided with a junction surface extending from the connecting part (211). The lower surface of the upper hub (111) and the upper surface of the lower hub (112) are separately provided with a matching surface matching the junction surface.

Abstract: The present disclosure describes a semiconductor structure and a method for forming the same. The semiconductor structure can include a substrate, a gate structure over the substrate, a layer of dielectric material over the gate structure, a source/drain (S/D) contact layer formed through and adjacent to the gate structure, and a trench conductor layer over and in contact with the S/D contact layer. The S/D contact layer can include a layer of platinum-group metallic material and a silicide layer formed between the substrate and the layer of platinum-group metallic material. A top width of a top portion of the layer of platinum-group metallic material can be greater than or substantially equal to a bottom width of a bottom portion of the layer of platinum-group metallic material.

Abstract: Provided are a manufacturing method for a hot-dip galvanized steel plate, and the hot-dip galvanized steel plate. The method comprises: hot rolling a slab into a steel plate, and pickling and cold rolling after coiling; performing continuous annealing, an annealing temperature being 840-870° C., and an annealing dew point being ?10-0° C.; cooling to 710-730° C. at a cooling rate of ?10° C./s, and then cooling to 220-320° C. at a cooling speed of ?50° C./s; then heating to 410-460° C. for heat preservation for 20-100 s; and galvanization to obtain the hot-dip galvanized steel plate, the chemical element composition thereof being: C: 0.17-0.21 wt %; Si: 1.2-1.7 wt %; Al: 0.02-0.05%; Mn: 1.60-2.1 wt %; N: ?0.008 wt %; and the reminder being Fe and impurities. The hot-dip galvanized steel plate of the present invention has a yield strength of 400-600 Mpa, a tensile strength of 730-900 Mpa, an elongation of 25-35%, and a hole expansion ratio of 35-60%.

Abstract: A display panel and a display device are provided. The display panel comprises a substrate, a plurality of sub-pixels, a plurality of data lines, a plurality of wiring structures and a plurality of pads. The substrate comprises a display area and a non-display area comprising a bending area. The sub-pixels and data lines are in the display area and electrically connected with each other. The wiring structures are in the bending area and electrically connected with the data lines. At least one wiring structure comprises a plurality of hollow patterns connected successively, each hollow pattern comprises a first conductive part and a second conductive part connected with each other. The plurality of pads are in the non-display area and located on a side of the plurality of wiring structures away from the display area and electrically connected with the plurality of wiring structures.

Abstract: An electro-hydraulic combination driven explosion-proof legged robot which includes a positive-pressure chamber body and a leg-foot assembly. The leg-foot assembly includes a hydraulic cylinder. A servo pump is arranged inside the chamber body. The hydraulic cylinder is located outside the chamber body. The chamber body has a chamber wall formed with oil inlet and outlet holes configured to intercommunicate the servo pump with the hydraulic cylinder. The leg-foot assembly includes a mounting bracket, a leg-foot forward-and-backward swing mechanism, a hip joint mechanism and a leg-foot mechanism. The leg-foot forward-and-backward swing mechanism and the hip joint mechanism are connected outside the chamber body through the mounting bracket. The hip joint mechanism is connected to the leg-foot mechanism and is able to drive the leg-foot mechanism to swing laterally inward and outward. The leg-foot forward-and-backward swing mechanism is connected to the hip joint mechanism in a coaxial transmission manner.

Abstract: Disclosed in the present disclosure is a method for smelting a high-temperature alloy with an ultrahigh N content in a VIM furnace, and a high-temperature alloy with an ultrahigh N content. The VIM-furnace smelting method comprises a smelting step performed in a VIM furnace, and a casting step, wherein at a later stage of smelting, a nitrogen-containing substance is added to the VIM furnace to adjust the content of N. In the present disclosure, by comprehensively improving the raw materials, the smelting process, etc., the yield of N during a VIM-furnace smelting process is increased and accurate control over the content of N in the high-temperature alloy is achieved, such that a target content of nitrogen can be guaranteed, and an ideal product can thus be obtained.

Abstract: Decoding methods and apparatuses, coding methods and apparatuses, devices and machine-readable storage media are provided. In an example of the decoding method, for a received bitstream, by parsing a motion information index value of a skip mode or a direct mode of a current coding unit, a candidate list of the skip mode or the direct mode may be constructed, and determining a prediction mode of the current coding unit, thus, motion compensation can be performed on the current coding unit according to motion information of the current coding unit acquired from the prediction mode of the current coding unit. The motion information index value of the skip mode or the direct mode of the current coding unit is obtained by parsing according to the number of motion vector angle prediction candidate motion information and the number of inter-prediction candidate historical motion vector prediction information.

Abstract: The present disclosure provides a crystalline form of calcium D-glycerate of formula and a method for preparing the crystalline form of calcium D-glycerate of formula (I). The crystalline form may be characterized by having X-ray powder diffraction pattern (XRPD) diffraction peaks (2? degrees) at 12.9°±0.2°, 20.9°±0.2°, and 31.7°±0.2°.

Abstract: Disclosed are a display panel and a display device. The display panel includes: a pixel unit, including a pixel circuit and a light-emitting element, the pixel circuit includes a driving transistor and a threshold compensation transistor; a first power line configured to supply a first power supply voltage to the pixel circuit; a blocker, electrically connected with the first power line; and a first conductive structure, connected with the gate electrode of the driving transistor, the first channel and the second channel of the threshold compensation transistor are connected by a conductive connection portion; an area of an orthographic projection of a portion of the blocker overlapping with the first conductive structure on the base substrate is larger than an area of an orthographic projection of a portion of the blocker overlapping with the conductive connection portion on the base substrate.

Abstract: The disclosure relates to the field of equipment for ice hockey training, and particularly discloses a single puck dribbling plate and a modular puck dribbling plate. A player can select the number and the combination mode of the single puck dribbling plates according to requirements, the training cost of the player is effectively reduced, and meanwhile different training requirements can be better met. The modular puck dribbling plate can be connected with terminal device, and data generated in the training process can be intelligently analyzed through the terminal device, so that the training effect of players can be better monitored. In addition, the modular puck dribbling plate has the advantages of being convenient to store and carry and the like.

Abstract: Techniques are described with respect to a system, method, and computer product for generating multi-media content based on linguistics. An associated method includes receiving a plurality of linguistic inputs and analyzing the plurality of linguistic inputs. The method further including generating multi-media content for presentation to a user based on the analyzing of the linguistic inputs.