[Q-e-developers] GW approximation: execution error in parallel mode

Paolo Giannozzi p.giannozzi at gmail.com
Thu Nov 16 09:56:19 CET 2017


I don't think GWW works with k-points

Paolo

On Wed, Nov 15, 2017 at 10:23 AM, Saiz, Fernan <
f.saiz-poyatos at imperial.ac.uk> wrote:

> Dear QE developers,
> I have downloaded via git the version 6.2 of Quantum Espresso to run
> GW-approximation calculations with a test case for silicon. My version is
> compiled with intel suite and mpi compilers as shown in the attached file
> qe_parallel_cx1.pbs.
>
> Unfortunately, I am getting the following MPI-related error at the end of
> the execution of the pw4gww.x command as you might see in the file
> ./pw4gww/log.out:
>
> *rank 19 in job 1  cx1-132-1-7_41723   caused collective abort of all
> ranks*
> *  exit status of rank 19: return code 0*
> *rank 16 in job 1  cx1-132-1-7_41723   caused collective abort of all
> ranks*
> *  exit status of rank 16: return code 0*
> *rank 10 in job 1  cx1-132-1-7_41723   caused collective abort of all
> ranks*
> *  exit status of rank 10: return code 0*
>
> As the serial version runs fine and I find the same error in a different
> unix machine, I wonder if this problem has to do with a possible bug in the
> code or, if I want to solve it, I should recompile the code with different
> options or compilers. A quick test on your side could shed some light on
> this problem.
>
> In addition, the execution of the serial version of the command gww.x
> seems to be hanging.
>
> Best regards,
>  Fernan Saiz
>  Department of chemistry
>  Imperial College London
>
> _______________________________________________
> Q-e-developers mailing list
> Q-e-developers at qe-forge.org
> http://qe-forge.org/mailman/listinfo/q-e-developers
>
>


-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20171116/25ea11b5/attachment.html>


More information about the developers mailing list