Abstract: GEMM (General Matrix Multiplication) is a fundamental operation in deep learning (DL), serving as the key computing kernel for neural network layers, such as convolutional and fully ...