[Pw_forum] scf calculation crashed for "diagonalization (ZHEGV*) failed".

GAO Zhe flux_ray12 at 163.com
Wed Jul 18 10:30:49 CEST 2012


If there is no model/structure problem, then you can try to set diagonalization='cg' in &electrons namelist, although this method will decrease the calculation speed due to the poor parallel performance.


--
GAO Zhe
CMC Lab, Materials Science & Engineering Department,
Seoul National University, South Korea
 

At 2012-07-18 13:30:15,"bamidele ibrahim" <bamideleibrahim at yahoo.com> wrote:

Dear all,
  I am running scf calculation  and relax with the input file below:

&control
  calculation = 'relax',
  prefix='MgSe'
  pseudo_dir='/home/adetunji/espresso-4.3/pslibrary.0.2.5/pz/PSEUDOPOTENTIALS',
  outdir='./',
/
&system
  ibrav=0,
  nat=8,
  celldm(1)= 11.32,
  ntyp=2,
  ecutwfc = 50.0,
/
&electrons
 mixing_beta = 0.9
 conv_thr = 1.0d-9
/
&ions
ion_dynamics ='bfgs'
/
ATOMIC_SPECIES
Mg 24.305  Mg.pz-n-kjpaw.UPF
Se 78.963  Se.pz-n-kjpaw.UPF
CELL_PARAMETERS
   0.000 0.49 0.49
   0.49  0.000 0.49
   0.49 0.49 0.000

ATOMIC_POSITIONS
Mg 0.00000000  0.00000000   0.00000000
Mg 0.00000000  0.50000000   0.50000000
Mg 0.50000000  0.00000000   0.50000000
Mg 0.50000000  0.50000000   0.00000000
Se 0.25000000  0.25000000   0.25000000
Se 0.75000000  0.75000000   0.25000000
Se 0.75000000  0.25000000   0.75000000
Se 0.25000000  0.75000000   0.75000000

K_POINTS (automatic)
6 6 6  1 1 1

In both cases, it crashes and produces this error:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from cdiaghg : error #        42
     diagonalization (ZHEGV*) failed
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Can anybody in the house tell me what i have done wrong.
                                                     


Adetunji Bamidele Ibrahim(PhD Student)
Department of physics,University of Agriculture,
Abeokuta, Ogun State,Nigeria.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20120718/e7c44ee2/attachment.html>


More information about the users mailing list