Abstract: A server and an assembly method thereof are disclosed. The server includes a chassis base, a first fan rack, a motherboard, a second fan rack, and a fan module. The first fan rack and the motherboard are both assembled on the chassis base, and the second fan rack is disposed on a side of the motherboard relatively close to the first fan rack, in which the first fan rack and the second fan rack form a receiving space, and the fan module is placed in the receiving space.

Abstract: An expansion card module includes a carrier, an expansion card disposed on the carrier, a heat-dissipation member disposed on the expansion card, and a latch assembly that is assembled to the carrier. The carrier includes a first end and a second end that is opposite to the first end. The second end has a pivotal portion. The expansion card has an electrical connection end that is arranged adjacent to the first end of the carrier. The expansion card is disposed between the carrier and the heat-dissipation member. The latch assembly is pivotally connected to the pivotal portion of the carrier.

Abstract: A method for detecting repairable boards requiring repair amongst many boards which may or may not require repair applies a board detection model based on training features of many sample repairable boards. The method obtains repair-relevant information of all the sample repairable boards, extracts predetermined features from the repair-relevant information, and analyzes the predetermined features to obtain the training features. The board detection model is established and trained based on the training features, and receives repair-relevant information of each repairable board to obtain a result of detection repairable board according to the board detection model. A device for detecting repairable boards is also provided.

Abstract: A method for detecting repairable boards requiring repair amongst many boards which may or may not require repair applies a board detection model based on training features of many sample repairable boards. The method obtains repair-relevant information of all the sample repairable boards, extracts predetermined features from the repair-relevant information, and analyzes the predetermined features to obtain the training features. The board detection model is established and trained based on the training features, and receives repair-relevant information of each repairable board to obtain a result of detection repairable board according to the board detection model. A device for detecting repairable boards is also provided.

Abstract: The present invention provides an aptamer binding to midkine and capable of forming a potential secondary structure represented by the formula (I): wherein X1, X2, X5 and X6 are the same or different and each is one or two nucleotides selected from the group consisting of A, G, C, U and T, or a bond, X1 and X6, and X2 and X5 each form a Watson-Crick base pairs, and X3 and X4 are the same or different and each is a nucleotide selected from A, G, C, U and T.

Abstract: The present invention provides an aptamer binding to midkine and capable of forming a potential secondary structure represented by the formula (I): wherein X1, X2, X5 and X6 are the same or different and each is one or two nucleotides selected from the group consisting of A, G, C, U and T, or a bond, X1 and X6, and X2 and X5 each form a Watson-Crick base pairs, and X3 and X4 are the same or different and each is a nucleotide selected from A, G, C, U and T.

Abstract: A volume sequential technique allows the production of prototypes from a liquid photopolymer precursor without requiring the CAD model to be decomposed into slices. Inverse Tomographic Construction (ITC) selectively cures the photopolymer in a vat without requiring it translating build platform.