[Pw_forum] performance of occupation='fixed' and ='smearing'

Gabriele Sclauzero sclauzer at sissa.it
Thu Sep 10 09:23:37 CEST 2009



shypirate wrote:
> Dear all:
> When I use pw.x, I find that performance of different occupations are 
> different. If I set occupation='smearing', it takes about 6 minutes for 
> one diagonalization step, but 24 minutes for occupation='fixed'. I have 
> set the same "nbnd" and "K_POINTS" for both. My question is, why is the 
> performance difference?

If it's like you say, the only possible difference that comes to my mind is in the number 
of iterations used in the Davidson algorithm for diagonalization (though I never thought 
it could depend on the occupations method). Please check for the keyword "iterations" 
(which is printed together with the number of iterations after each diagonalization step):
 > grep iterations yourfile
and let me know the output.

GS

> Best Wishes,
> Y.P.Wang
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> 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