For running openmp programs run commands:- g++ filename.cpp -fopenmp and ./a.exe [for windows users] or ./a.out [for linux users]
Make sure MinGW is installed with pthreads.
If you still get errors try running: mingw-get upgrade --recursive "gcc<4.7." "gcc-g++<4.7."
To run CUDA programs on Collab, follow these steps: Go to Google Collab Create a new Notebook(.ipynb file). Click on Runtime and change runtime type to GPU. Now run !pip install git+https://github.com/afnan47/cuda.git in a cell. On a new cell run %load_ext nvcc_plugin Test the following code %%cu #include int main(){ std::cout << "Hello World\n"; return 0; } Remember to add %%cu before writing the C++ code for every CUDA program. CUDA is now set. Click here for detailed steps