make [optimization-seq|optimization-openmp|optimization-mpi]
For mpi you need to change the value of BINROOT to the path of mpic++
- Sequentcial :
./optimization-seq [function precision] - Parallel (OMP) :
./optimization-openmp [function precision] - Parallel (MPI) :
mpirun -host host1,host2 -n 4 ./optimization-mpi [function precision]
On MPI parallel version, the result is not guaranteed if the number of processes (-n) is not equal to 4
If you run without parameter, you have access of help.