Parallel Programming with Intel MPI

|Chapter 1: Introduction
|Chapter 2|Chapter 3|Chapter 4|

Download and install Intel MPI

Please check the hardware, software and other development tools requirement for your need on intel's webpage.

For my system which has 16 GB RAM, and a 2.1GHz quad core Intel Core i-5. Install the following tools. form the corresponding website.

1. Microsoft Visual Studio 2019 (Community Edition). 

    We will only select Desktop development with C++ set of tools in our installation.

2. Intel OneAPI Base toolkit

3. Intel OneAPI HPC Toolkit 

We need to install both base and hpc toolkit from intel oneapi website . One has to have or create an account  be able to download . Selet appropriate OS, installer type etc. on the webpage. The appropriate download button will show up. Just click it.

The HPC toolkit provides a set of compiler and libraries.  DPC++/C++ compilers and the intel MPI Libraries are required for this tutorial others are optional. 

After download is complete the installations are straightforward. Just keep clicking next or appropriate options provided on the prompt.You can choose installation path on your system while installing. A default installation path could be in - "C:\Program Files (x86)\Intel\oneAPI". To be able to compile and run using any intel compilers we need to run follwoing command on our command prompt.

       
         "C:\Program Files (x86)\Intel\oneAPI\setvars.bat" intel64
         
   

This will set all environment variables and paths needed for compiling and running a program with MPI.

MPI: Compile and Run your first program.

Let us write our first MPI program. Here is mphello.c Hello World program program using MPI.

       
#include &ltmpi.h&gt
#include &ltstdio&gt
/*mphello.c: Hello World program program using MPI.*/

int main(int argc, char** argv) {
    // Initialize the MPI environment
    MPI_Init(NULL, NULL);

    // Get the number of processes
    int world_size;
    MPI_Comm_size(MPI_COMM_WORLD, &world_size);

    // Get the rank of the process
    int world_rank;
    MPI_Comm_rank(MPI_COMM_WORLD, &world_rank);

    // Get the name of the processor
    char processor_name[MPI_MAX_PROCESSOR_NAME];
    int name_len;
    MPI_Get_processor_name(processor_name, &name_len);

    // Print off a hello world message
    printf("Hello world from processor %s, rank %d out of %d processors\n",
           processor_name, world_rank, world_size);

    // Finalize the MPI environment.
    MPI_Finalize();
}
       
 

Note: It will help to go through the explanation of the above code in Dr. Wes Kendall's original website. 

To compile the program:

       
mpicc mphello.c -o mphello
   

This will create a mphello.obj and mphello.exe in your local directory. To run the program:

       
mpiexec mphello         
    
    output:
    Hello world from processor COMPUTER-USER, rank 1 out of 2 processors
    Hello world from processor COMPUTER-USER, rank 0 out of 2 processors
         
   

|Chapter 1|Chapter 2|Chapter 3|Chapter 4|

Example : Trapezium Rule

we can use the trapezoidal rule to approximate the area between the graph of a function, y = f(x), two vertical lines, and the x-axis. The basic idea is to divide the interval on the x-axis into n equal subintervals. Then we approximate the area lying between the graph and each subinterval by a trapezoid whose base is the subinterval, whose vertical sides are the vertical lines through the endpoints of the subinterval, and whose fourth side is the secant line joining the points where the vertical lines cross the graph. See Figure 3.4. If the endpoints of the subinterval are xi and xi+1, then the length of the subinterval is h = xi+1 − xi . Also, if the lengths of the two vertical segments are f(xi) and f(xi+1), then the area of the trapezoid is Area of one trapezoid = h/2 *[ f(xi) + f(xi+1)]

Here is the serial implementation

         
  double Trap(double left ,double right,int trap_count  ) {
  double estimate, x;
  int i;
  double base_len = (right -left)/trap_count;
  estimate = (func(left) + func(right))/2.0;
  for (i=1;i<=trap_count;i++) {
      x = left + i*base_len;
      estimate += func(x);
  }
  estimate = estimate*base_len;
  return estimate;
 }
  

|Chapter 1|Chapter 2|Chapter 3|Chapter 4|

Example : Galerkin Method

|Chapter 1|Chapter 2|Chapter 3|Chapter 4|

Advanced level routines

More Resources:

mpitutorial.com