[Pw_forum] parallelization issues

Gabriele Sclauzero sclauzer at sissa.it
Thu Mar 5 09:45:04 CET 2009

Mahmoud Payami wrote:
> Dear ALL,
> In sec. of "parallelization issues" of UG, is mentioned that: "As a 
> general rule, image parallelization may give good scaling..."

These is relevant to you only if you are performing NEB calculations.

> Could anybody please explain what exactly it means in terms of say, 
> static linking, using more npools, or so.

I think it has nothing to do with static linking and in general with any compilation 
issue. Once you correctly compiled a parallel version, you're free to chose which kind of 
parallelization to eploit, depending of your type of system and calculation.
If you're doing NEB, then the image parallelization is the more efficients since different 
images have to "speak" very few one to the other, while if not doing NEB you can still use 
the other levels of parallelization described in the user guide. They do not scale as well 
as image parallelization, but still do a good job to speed up your code or let you study 
very big systems.



> Bests,
>            Mahmoud Payami
> Physics Group, AEOI.
> ------------------------------------------------------------------------
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum


o ------------------------------------------------ o
| Gabriele Sclauzero, PhD Student                  |
| c/o:   SISSA & CNR-INFM Democritos,              |
|        via Beirut 2-4, 34014 Trieste (Italy)     |
| email: sclauzer at sissa.it                         |
| phone: +39 040 3787 511                          |
| skype: gurlonotturno                             |
o ------------------------------------------------ o

More information about the users mailing list