[Pw_forum] phonon calculation in Ni on 7x7x7 mesh problem.

[Andrea Dal Corso]

Open Module/parameters.f90 and increase npk. For this calculation you
need at least 27648 x 2 but maybe more for other q points.
Then recompile.

HTH,

Andrea

On Tue, 2015-07-14 at 11:24 -0400, German Samolyuk wrote:
> Dear QE developers, users,
> 
> I'm trying to calculate phonon in Ni with ultrasoft
> pspot Ni.pbe-nd-rrkjus.UPF with 7x7x7 mesh.
> It perfectly works for nxnxn with n <= 6 but fail with larger n starting
> from 7.
> for  q =    0.5714286  -0.2857143   0.8571429
> 
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      Error in routine set_kup_and_kdw (27648):
>      too many k points
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> 
>      stopping ...
> 
> I run
> 
> mpirun ./ph.x -npool 64 < ni.ph.in > ni.ph.out
> 
> 
> with ni.ph.in
> 
> Electron-phonon coefficients for Ni
>  &inputph
>  tr2_ph=1.0d-12,
>  prefix='Ni',
>  amass(1)=58.70,
>  outdir='.',
>  fildyn='ni.dyn',
>  ldisp=.true.
>  nq1=7, nq2=7, nq3=7
> /
> 
> 
> 
> 
> 5.1 QE version
> 
> intel based cluster 16 cores per node. QE espresso compiled with intel
> compiler.
> 
> 
> 
> What can cause this problem?
> 
> 
> Thank you,
> 
> German,
> 
> German Samolyuk, ORNL, USA
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum

-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Bonomea 265              Fax. 0039-040-3787249
I-34136 Trieste (Italy)             e-mail: dalcorso at sissa.it