The ABINIT Users Mailing List ( CLOSED )

[paul.boulanger@umontreal.ca]

Hi,

We are (group of Michel Côté in Montreal) in the process of testing the 
modifications of X. Gonze in version 5.4.4. Are preliminary results shows 
that the number of communicators and groups does not grow with the number of 
datasets, a clear improvement on the older versions. But, we still need to 
specify MPI_GROUP_MAX and MPI_COMM_MAX greater then the number of K-points 
(we are using twice the number of K-points for these variables).

We are currently looking to fix this last (hopefully) issue. Maybe someone 
more knowledgeable will find a workaround with your platform.

Cheers,

Paul