Parallel Programming And Optimization Based On Multi-core Processors

With the rapid development of parallel computer technology, especially multi-core technology,the programmer will design parallel programs. OpenMP has became the standard of shared memory structure programming because of its easy-to-use features.This paper first introduces parallel computing architecture,memory access model,and compare the current popularity of several parallel programming model. Then focus on OpenMP programming model and analysis the fork-join parallel programming design mode and SPMD mode of OpenMP. Based on the analysis of the factors that affect the performance of parallel algorithms, we proposed some programming optimization method of OpenMP. Then the experiments of parallelized ant colony algorithm for the TSP problem shown the improved parallel algorithm can shorten the solution time and improve the efficiency of the implementation.Then, using the MPI+OpenMP hybrid parallel programming model on a cluster system, and further research on some of the issues that need attention in the hybird parallel programming and some optimization methods. We proved this hybrid parallel programming model can be a good way to the advantages of multi-core processors by matrix multiplication experiment At last, the paper introduces the basic principles of synthetic aperture radar, radar imaging algorithm with sub-aperture imaging technology, and give the parallel process of the NCS algorithm. According to demand analysis, we choose FPGA and DSP for real-time imaging system, and proved that the NCS real-time image processing system can meet the real-time demand by the theoretical analysis of computational speed, storage capacity and communications bandwidth.