Update for running with -nk (or -npools) option.
Thank you, @MouadBik, for pointing out and helping us to fix this issue.
To optimize the calculation of electron-phonon matrix elements, we can run the program with the specified option:
$mpirun -np 48 ph_mat.x -nk 48 <ph.in> ph.out
It can be faster than the case of without "-nk 48"
Note: The number of k-points in scf.in should be a multiple of the number of MPI tasks.
Contributors
MouadBik