第一题

#include #define  NUMBER   4 void main( ) {    int i=1; float f0,dt,dx,dz; int mx,mz,ntmax,pml,sou_x,sou_z,rec_start,rec_end,rec_int,rec_depth; char shotID[5],record_pp[30],vp_name[30],rou_name[30],mig_name[30],mig_pyt_name[30],mig_udlf_name[40],shuoming[30]; FILE *fp,*f; /*&不合法变量*/     printf("** This program is to show the block file input & output** "); fp=fopen(

进行相关运算时必须要是用的一些步骤,可以解决相关算法的一些问题(Related operations must be used in a number of steps can be solved a number of issues related to algorithm)

说明：  并行计算实验，多线程计算pi源代码，附有说明。(Parallel computing experiment, multithreaded to calculate pi source code annotated.)

LU算法的MPI串行实现，矩阵维数无要求，可在曙光4000A并行机运行(The LU algorithm MPI serial implementation matrix dimension requirements Dawning 4000A parallel machine running)

使用二进制PSO解决0-1背包问题的MATLAB程序(The use of PSO to solve 0-1 knapsack problem)

MPICH是MPI的重要研究,提供了一系列的接口函数,为并行计算的实现提供了编程环境.(MPICH is an important MPI Research, has provided a series of interface functions, for the realization of parallel computing provides a programming environment.)

并行计算实验，wordcount算法Mapreduce实现。(Parallel computing the experiment, wordcount algorithm Mapreduce achieve.)

《GPU高性能编程CUDA实战》书中第5章源代码(Chapter 5 source code of the GPU high-performance programming CUDA real " book)

应用背景只是用来学习OPENMP编程的应用技巧，同时对初学者有所帮助，不过程序比较简单，谨慎学习下载，同时，希望学习openmp的朋友能够多多交流一下，相互学习帮助关键技术关键技术就是基于OPENMP的一个矩阵相乘的并行实现，然后使用了分块，用静态分块进行各线程并行处理，所用时间短，效率高，适合学一下

cuda资料，对于并行编程很有帮助！该资料描述了cuda的基本信息，以及重要的步骤，让您逐渐上手，轻松掌握cuda并行编程。(cuda information useful for parallel programming! The information describes the cuda s basic information, as well as an important step, so that gradually you get started, easy to master parallel programming cuda.)