The aim of this project is to perform convolution in FPGA. Firstly, the input and the convolution weight is loaded to buffer. Then, the control unit send the read address to the buffer. As a result, the computation unit is just only multiply-accumulate until the computation finish