[Pw_forum] GIPAW 5.3.0 "problems computing cholesky"

[Ary Ferreira]

Dear Jennifer,

You could try to decrease the convergence threshold for self-consistency
(input variable conv_thr). Try 1.0D-10 and if you have problems with its
convergence there are some options related to the mixing scheme. Maybe you
will have to work on this PP for Erbium from SSSP (RE Wentzcovitch) before
use it in your system.

Ary Ferreira,

CNPq postdoctoral fellow
TU München


On Thu, Apr 21, 2016 at 10:26 PM, Elward, Jennifer M CTR USARMY RDECOM ARL
(US) <jennifer.m.elward.ctr at mail.mil> wrote:

> Hello,
>
> I am trying to use the GIPAW module with QE to compute magnetic
> susceptibility and NMR parameters. I am using QE v5.3.0 and GIPAW v5.3.0.
> My system has 120 atoms and the input file for GIPAW is copied below.
>
> &inputgipaw
>         job = 'nmr'
>         prefix = 'al2o3_er'
>         isolve = 0
>         tmp_dir= './'
>         iverbosity = 1
>         q_gipaw = 0.01
>         spline_ps = .true.
>         use_nmr_macroscopic_shape = .false.
> /
>
> The error I receive is the following,
>
> Error in routine  cdiaghg (1431):
>       problems computing cholesky
>
> It seems as though this is a common problem and I have seen the potential
> solutions which included adding -ndiag 1 to the execution, changing isolve
> from 0 to 1 or modifying q_gipaw from 0.01 to a higher number. I have tried
> these solutions as well but the error I receive above remains.
>
> I was wondering if there were any additional solutions to this problem or
> other parameters to change to make the susceptibility calculation work?
> Input and output files are attached.
>
> Thanks very much!
>
> Jennifer Elward, PhD
> Army Research Laboratory
>
>
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
http://lattes.cnpq.br/8221674673413336
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160422/d5b1bfca/attachment.html>