I have installed the paralution plugin for OpenFOAM and I am currently working on a node with two 10-core CPUs and 2 GPUs. Normally, when I want to run the icoFoam case without using a GPU, I use the command ‘mpirun -np 20 -parallel icoFoam’ to run the case in parallel.
Now i want to accelerate it using a GPU. Is it sufficient to just modify the fvSolution file according to the Paralution tutorial? Or do I have to change the -np option of mpirun also?