Dear Haowei Peng,
in my experience this error may happen with GGA functionals when you push
the scf
threshold very low (let's say conv_thr below 10^(-10) ). Since you're hiding
some relevant
information of your input file (such as PP files, atomic coordinates and
k-points) I
cannot be sure if this is causing the problem.
In your case conv_thr = 1.0d-8, which is not that low, so there may be other
problems,
such as a wrong set of coordinates on input, a bad pseudopotential, ...
Maybe, you can try to increase the cutoffs.
HTH
GS
Haowei Peng wrote:
> Hi, all,
>
> I am running a scf calclation, but it fails several times due to the
> same reason.
> However, the relax calculation using similar parameters and input file
> goes well.
> the input file looks:
>
> &control
> calculation='scf'
> restart_mode='from_scratch',
> nstep=50,
> pseudo_dir = '/home/haowei/tools/pwscf/pseudo/',
> outdir='./results/'
> prefix='pwscf'
> etot_conv_thr = 1.0d-5,
> forc_conv_thr = 1.0d-4,
> /
> &system
> ibrav = 0,
> celldm(1) = 6.501000,
> nspin = 1
> nat = 28,
> ntyp = 3,
> ecutwfc = 25,
> ecutrho = 200,
> occupations='smearing', smearing='gauss', degauss=0.001,
> lda_plus_u=.false. Hubbard_U(1)=1.d-8
> /
> &electrons
> mixing_mode = 'plain'
> mixing_beta = 0.3
> conv_thr = 1.0d-8
> /
> .....
>
> and this is the crash file
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> task # 4
> from cdiaghg : error # 608
> diagonalization (ZHEGV*) failed
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
> thanks a lot
>
> Best,
> --
> Haowei Peng
> Northwestern University
> Department of Physics & Astronomy
> 2145 Sheridan Rd.
> Evanston, IL 60208-3112
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
--
o ------------------------------------------------ o
| Gabriele Sclauzero, PhD Student |
| c/o: SISSA & CNR-INFM Democritos, |
| via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511 |
| skype: gurlonotturno |
o ------------------------------------------------ o
