How to use OpenMP to parallelize C++ using Visual studio

Parallel programming is the process of breaking down a large task into smaller sub-tasks that can be executed simultaneously, thus utilizing the available computing resources more efficiently. OpenMP is a widely used API for parallel programming in C++. It allows developers to write parallel code easily and efficiently by adding simple compiler directives to their existing code.

The OpenMP C++ application program interface lets us write applications that effectively use multiple processors. In this tutorial, we will cover how to run a parallel code in Visual C++, which supports the OpenMP 2.0 standard.

Outline

1. Configure C++ project
   • Create a new C++ project
   • Enable OpenMP support
2. Run code
   • Add code for parallel loop and Run

Source code: https://github.com/tuttelikz/notes/tree/main/openmp-vs/OpenMPApp

---

Configure CPP project

Create a new cpp project

Open Visual Studio > Create a new project > Set "C++", "Windows" and "Console" filters to find the template > click Next

Enable OpenMP support

Navigate to Project > Properties

Then C/C++ > Language > Open MP Support > Choose "Yes" > Apply > OK

Importanly, make sure precompiled headers are off by Precompiled Headers > Precompiled Header > Not Using Precompiled Headers > Apply > OK

---

Run code

Add code to OpenMPApp.cpp and Run

To test OpenMP functionality in C++ project, we defined a simple sum function which calculates result using conventional serial loop and parallel loop. Add the following code to OpenMPApp.cpp

#include <chrono> 
#include <iostream> 
#include <omp.h>

using namespace std;

// Serial for loop
int sum_serial(int n)
{
    int sum = 0;
    for (int i = 0; i <= n; ++i) {
        sum += i;
    }
    cout << "Serial result: " << sum
        << endl;

    return sum;
}

// Parallel for loop
int sum_parallel(int n)
{
    int sum = 0;
#pragma omp parallel for reduction(+ : sum) 
    for (int i = 0; i <= n; ++i) {
        sum += i;
    }
    cout << "Parallel result: " << sum
        << endl;

    return sum;
}

int main()
{
    int num_threads = 8;
    omp_set_num_threads(num_threads);

    const int n = 100000000;

    auto start_time
        = chrono::high_resolution_clock::now();
    int result_serial = sum_serial(n);
    auto end_time
        = chrono::high_resolution_clock::now();
    chrono::duration<double> serial_duration = end_time - start_time;

    start_time = chrono::high_resolution_clock::now();
    int result_parallel = sum_parallel(n);
    end_time = chrono::high_resolution_clock::now();
    chrono::duration<double> parallel_duration = end_time - start_time;

    cout << "Serial time elapsed: "
        << serial_duration.count() << " seconds"
        << endl;
    cout << "Parallel time elapsed: "
        << parallel_duration.count() << " seconds"
        << endl;
    cout << "Speedup for " << num_threads << " threads: "
        << serial_duration.count()
        / parallel_duration.count()
        << endl;

    return 0;
}

Press Start debugging to see the result

Thanks for reading, hope this post was useful. Stay tuned, more to come 🙂

Alternate URL: https://github.com/tuttelikz/notes/tree/main/openmp-vs