Research And Implementation Of Parallel Processing System Based On FPGA

FPGA is a field programmable gate array,which can simulate various parallel operation mechanisms of hardware.The system based on FPGA usually has the flexibility of programming and high processing speed.Because of the unique hardware programmable characteristics,FPGA requires specially designed computing system to achieve the maximum utilization of hardware resources in various fields of computing problems.In this paper,the algorithm based on GTM model is proposed for NP-complete problems such as equipartition problem and 0-1 knapsack problem,and the possible results of the problem can be obtained in multiple times.In this paper,an Efficient parallel computing component,EP(Efficient Parallel)is designed,and the whole parallel processing system is designed and implemented based on this component.Traditional processing system uses serial processing mode to solve the above problems,which requires a lot of computing time.However,this system can rely on EP(Efficient Parallel)to accelerate the above problems in parallel and obtain more computing capacity than traditional processing system.The main contents of this paper are as follows:(1)a parallel processing system is designed and implemented based on GTM model.The system includes controller,memory,system I/O and EP arithmetic unit.The system mainly relies on EP operation array and address converter for parallel operation to accelerate the problem solving.(2)In order to solve the equipartition problem and 0-1 knapsack problem,the corresponding algorithm is proposed based on the parallel processing system.(3)Solve the specific 10 variables equalization problem,and realize FPGA board-level debugging,and perform visual output on LCD screen.(4)Solve the specific 10-variable 0-1 knapsack problem,and realize FPGA board-level debugging,and perform visual output on LCD screen.