[Pw_forum] ZGEMM or MATMUL for MPI environment

[Paolo Giannozzi]

If the matrix S is not distributed (and if you have a copy of it on each
processor), MATMUL (evc, S) will be distributed exactly in the same way as
evc

Paol

On Wed, Aug 31, 2016 at 9:16 PM, Ilya Ryabinkin <igryabinkin at gmail.com>
wrote:

> Colleagues:
> I'm wondering what should I use to get the matrix product evc*S, where
>    evc  = evc(npw, nbnd)
> is a collection of MOs and
>    S = S(nbnd, nbnd)
> if my code is executed on several MPI nodes?
>
> I see that for for nproc >1
>
> npw is actually npw/nproc, which brings a question what is happening
> in the MATMUL(evc, S)
>
>
> Ilya
>
> --
> *******************************************************
>                     Ilya Ryabinkin
>                  Postdoctoral Scholar
>           Physical and Environmental Sciences
>            University of Toronto Scarborough
>   http://www.utsc.utoronto.ca/~aizmaylov/Members.html
> *******************************************************
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160831/bcb5b002/attachment.html>