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: Improved techniques for testing an application to identify bugs. An API request body, which includes input data, is transmitted to a service to exercise the service. An error type response is received from the service, where the response indicates how the service handled the input data. The response is then used to determine an error type response coverage of the service. The coverage is then expanded by repeatedly performing a number of operations until a threshold metric is satisfied. For instance, in response to learning how previously-used input data impacted the coverage, new input data is generated. This new input data is designed to trigger new types of responses from the service. The new input is sent to the service, and a new error type response is received. These processes are repeated in an effort to expand the error type response coverage of the remote service.

Abstract: Improved techniques for testing an application to identify bugs. An API request body, which includes input data, is transmitted to a service to exercise the service. An error type response is received from the service, where the response indicates how the service handled the input data. The response is then used to determine an error type response coverage of the service. The coverage is then expanded by repeatedly performing a number of operations until a threshold metric is satisfied. For instance, in response to learning how previously-used input data impacted the coverage, new input data is generated. This new input data is designed to trigger new types of responses from the service. The new input is sent to the service, and a new error type response is received. These processes are repeated in an effort to expand the error type response coverage of the remote service.

Abstract: A chip includes a substrate; macros placed on the substrate, which has a placement region being divided into sub-regions according to locations of the macros; and one or more vertical power stripes (VPSs) disposed in each sub-region. At least one VPS is not aligned with the VPSs of an adjacent higher or lower sub-region.

Abstract: A chip includes a substrate; macros placed on the substrate, which has a placement region being divided into sub-regions according to locations of the macros; and one or more vertical power stripes (VPSs) disposed in each sub-region. At least one VPS is not aligned with the VPSs of an adjacent higher or lower sub-region.

Abstract: A chip includes a substrate; macros placed on the substrate, which has a placement region being divided into sub-regions according to locations of the macros; and one or more vertical power stripes (VPSs) disposed in each sub-region. At least one VPS is not aligned with the VPSs of an adjacent higher or lower sub-region.

Abstract: A chip includes a substrate; macros placed on the substrate, which has a placement region being divided into sub-regions according to locations of the macros; and one or more vertical power stripes (VPSs) disposed in each sub-region. At least one VPS is not aligned with the VPSs of an adjacent higher or lower sub-region.

Abstract: A tread-replaceable environmental-protection wheel tire includes a tire body and a tread layer. The tire body has two sides each having an end forming a bead section. The bead sections include reinforcement members embedded therein. The bead sections are positionable against a wheel rim. The tire body has a top face that forms fitting slots. The top face of the tire body has two side shoulders each forming a retention slot. The tread layer has a top surface forming tread patterns. The tread layer has a bottom surface forming fitting ribs raised therefrom. The fitting ribs are receivable into the fitting slots of the tire body. The tread layer has two side edges each forming a retention section. The retention sections are fixable in the retention slots of the tire body. As such, the applicability of wheel tire is expanded and the costs for replacing wheel tire are reduced.

Abstract: A light emitting device, includes a substrate; a first semiconductor layer on the substrate; an active layer on the first semiconductor layer; a second semiconductor layer on the active layer; a transparent conductive layer on the second semiconductor layer; and a plurality of pillar structures with a hollow structure in the portion surface of the first semiconductor layer, thereby, the light extraction efficiency of the light emitting device can be improved due to the pillar structures with a hollow structure.

Abstract: A light emitting device, includes a substrate; a first semiconductor layer on the substrate; an active layer on the first semiconductor layer; a second semiconductor layer on the active layer; a transparent conductive layer on the second semiconductor layer; and a plurality of pillar structures with a hollow structure in the portion surface of the first semiconductor layer, thereby, the light extraction efficiency of the light emitting device can be improved due to the pillar structures with a hollow structure.

Abstract: A method for accessing data is disclosed, which method is for accessing data to a plurality of storage media through an electronic communication device, wherein the storage media are electrically connected to a gateway, the method including the steps of: the electronic communication device being connected to the gateway in a wireless manner; the electronic communication device selecting a file type of the data to be accessed; the electronic communication device sending the file type to the gateway; the gateway searching, according to the file type, for the data in the storage media, which data matches with data to be accessed; and the gateway sending the matching data to the electronic communication device.