Research of Artificial Neural Network Arithmetic Based on FPGA

FPGA implementation of Artificial Neural Networks(ANNs)has been an approach of actual real-time neural network.This paper systematically reviews the progress that has been made in this research area over a decade,and lists the typical technology issues,shows detailed data analysis and newest research results,Simultaneously different implementation techniques and design are discussed,then existing problems and realizition constrains can be presented as well as improvement solutions,the development direction and potential of FPGA implement of ANN are proposed.Eventually future developments are expected.