[Wien] ELPA

[Laurence Marks]

For details everything depends upon compilation. My experience is that ELPA
is faster except for very small sizes (e.g. bulk MgO) where the mpi version
can crash. I always use it in both lapw1 & lapw1_mpi. Best is for you to
test on your system.

Note that ELPA (and all of scalapack/lapack) work well with mpi, openmp and
hybrid (both). Hybrid mode with 2 openmp & many mpi can be very useful if
the memory/core ratio limits large calculations.
___
Professor Laurence Marks (Laurie)
Department of Materials Science and Engineering, Northwestern University
www.numis.northwestern.edu
"Research is to see what everybody else has seen, and to think what nobody
else has thought" Albert Szent-Györgyi

On Sat, Aug 24, 2024, 17:36 Arthur H. Edwards <edwardsah3 at fastmail.fm>
wrote:

> At approximately what dimension does ELPA become useful. I plan on doing
> single point, total energy calculations to calculate hyperfine parameters
> for ~216 atom cells of SiC, GaN/AlN with defects having symmetry of C3v, or
> C2v. I will be computiing on an MPI system with between hundreds and
> thousands of nodes, each with between 16 and 32 processors. We use openmpi
> and openblas.
>
>
> Thanks,
>
> Art Edwards
> --
>   Arthur H. Edwards
>   edwardsah3 at fastmail.fm
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20240825/a25724aa/attachment.htm>