[Pw_forum] MPI

[Mike Klymenko]

 Dear QE community,

I am beginner in QE.  I do successfully pw-scf and pw-hscf computations for
silicon. I have already installed MPI libraries. What is the easiest way
for beginner to start efficient parallelization of computations?

I have read the user guide and find an example with many parameters formed
a kind of hierarchy:

mpirun -np 4096 ./neb.x -nimage 8 -npool 2 -ntg 4 -ndiag 144 -input my.input

Do I really need  to specify all of then for scf? My computer has 32 CPUs.
I know that practice makes perfect, but could I get some knowledge a priory
here?

Thanks in advance!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20131009/30acd2f4/attachment.html>