The ABINIT Users Mailing List ( CLOSED )

[Marc Torrent <marc.torrent@cea.fr>]

Hi Joe and Emmanuel,

It is obvious that you just found the source of a recurrent issue in paral_kgb
computations.
We were here (at the cea) looking for the origin of our convergency problems
without having identified it; it is a difficult task because Abinit converges
but slowly...
We now need to correct the error.
We also need to check the validity of dilatmx parameter with paral_kgb
(generally connected to ecutsm).

I'll transmit the problem to the people actually working on parallelism...
hopefully that they will have to time...
Is you want to take part to the bug resolution and investigate further, your
help is welcomed.

Marc