[Pw_forum] BUG in NEB calculations with tefield=true

[Lorenzo Paulatto]

I've noticed with qe 6.2 that unless I manually set disk_io="medium" the 
NEB fails at the second step, on my computer, because it cannot read the 
wavefunctions of image 2 from the previous step.

If you had some stale data from a previous run, that may be read and 
cause problems, although I'm not sure this is the cause of your problem, 
it definitely is a bug in NEB.

Can you try to:
1. clean properly the temporary directory
2. rerun your neb simulation with disk_io="medium"

Hope this helps,
let me know if this solves the problem for you, we can look for a fix in 
the code.

cheers

p.s.

On a side note, using celldm(1) to convert Bohr to Angstrom is 
deprecated, do not define celldm(1) and use
   CELL_PARAMETERS angstrom
instead.



On 10/11/17 19:20, Paolo Giannozzi wrote:
> Please add option verbosity='high' to &control .. / and observe the 
> various force terms computed at the end of the first scf calculation for 
> the first image (tail -f tmp/neb_1/PW.out).
> 
> Paolo
> 
> On Tue, Nov 7, 2017 at 4:20 PM, Maxim Skripnik 
> <maxim.skripnik at uni-konstanz.de <mailto:maxim.skripnik at uni-konstanz.de>> 
> wrote:
> 
>     The smallest example I have here is a molecule with 58 atoms. It
>     takes 7 hours on 16 cores and produces the described behavior. But
>     one doesn't have to wait until NEB has converged. You can see the
>     NaN and ***** in the .path file already after the first NEB step,
>     which takes ~15 minutes on 16 cores.
> 
>     See attachment for the input file.
> 
>     _______________________________________________
>     Pw_forum mailing list
>     Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>     http://pwscf.org/mailman/listinfo/pw_forum
>     <http://pwscf.org/mailman/listinfo/pw_forum>
> 
> 
> 
> 
> -- 
> 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
> 
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
> 

-- 
Lorenzo Paulatto - Paris