Design and Implementation of Intelligent Hardware of Neural Network Based on FPGA

Neural networks, with its inherent parallel nature, is not suitable to realize in microprocessor, however, FPGA is widely applied in the hardware realization of neural network, for its parallel nature, abundant embedded hardware core unit with multiplying and adding function, and memory resources. According to the hardware circuit design, the neural network system is divided into several modules, and then the hardware implementation of each module is designed. The hardware circuit is realized by Verilog language, and the simulation test is carried out in Quartus II. According to the simulation results, whether the expected results are achieved is analyzed.