ABINIT Discussion Forums

OK, thanks. I see now.

I am not using mpirun, just evoking abinit from the command line (see earlier post in this thread).

Thank you for the suggestion. However, the problem persists.

>./configure --prefix=/home/fs03/ehs73/bin/abinit-7.6.2
> make clean
> make multi multi_nprocs=8
> make install

Followed by
abinit <input.files>& log

still only runs on one processor.

Hello, I'm compiling abinit 7.6.2 to run on a cluster. We have 8 processors per node, so I used >make multi_nprocs=8 when I was compiling. Compilation ran without errors. When I run abinit, I do not use mpi to handle the job. Rather, I log in to a compute node (which has 8 processors) and run abinit...

Are you generating the identical kmeshes, i.e. are the kpoints in the first (primitive) BZ in the same points? If not, perhaps your difference is one of numerical precision and convergence.

OK, thanks. I am just a beginner with abinit but hadn't encountered cutoffs of more than ~30 or so in the tutorial or forums. Of course, the examples I happened to see were not for first-row elements.

I'm doing a simple ecut energy convergence of CsPbF3. I'm using HGH pseudopotentials and the zero-T, zero-P structure. My energies don't converge to within 10 meV/f.u. until my ecut is over 150! I have not set the energies to eV, so they should still be in Ha, correct? Is this value of 150 OK?? Than...