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 charging device for an electric vehicle includes a network connection unit, a display light-panel, and a processor. The network connection unit is coupled to a server through a network and receives an indication of an operation of a charging procedure performed by a user device corresponding to the electric vehicle via the network. The display light-panel has an optical light-guiding component and a backlight module, wherein the optical light-guiding component has a fixed pattern printed thereon and the backlight module includes light sources, and the optical light-guiding component and the light sources are respectively disposed to operate the display light-panel in at least one illumination mode. The processor respectively changes the illumination mode of the display light-panel in response to the received indication, such that the display light-panel changes the light sources to generate an interaction display effect corresponding to the operation of the charging procedure.

Abstract: A charging device for an electric vehicle includes a network connection unit, a display light-panel, and a processor. The network connection unit is coupled to a server through a network and receives an indication of an operation of a charging procedure performed by a user device corresponding to the electric vehicle via the network. The display light-panel has an optical light-guiding component and a backlight module, wherein the optical light-guiding component has a fixed pattern printed thereon and the backlight module includes light sources, and the optical light-guiding component and the light sources are respectively disposed to operate the display light-panel in at least one illumination mode. The processor respectively changes the illumination mode of the display light-panel in response to the received indication, such that the display light-panel changes the light sources to generate an interaction display effect corresponding to the operation of the charging procedure.

Abstract: An expansion structure for compact computers to install expansion circuit boards therein is provided. This expansion structure is small in dimension to save space in a compact computer and also allows easy installation of the expansion circuit boards. The expansion structure includes a base board on which an expansion circuit board is mounted, an enclosure having at least one engaging slot for accommodating the base board therein, and a main board which is mounted in the enclosure and formed with an electrical connector. The electrical connector is mounted in a connector receptacle provided on one edge of the engaging slot for the main board to have an electrical connection with the expansion circuit board. In installation the base board is inserted into the engaging slot allowing electrical connection pins on the expansion circuit board to come into contact with the electrical connector. After that securing means is used to secure the expansion circuit board in position in the engaging slot.