[Pw_forum] Very long time of computation

Paolo Giannozzi giannozz at nest.sns.it
Mon Oct 17 09:30:26 CEST 2005


On Sunday 16 October 2005 16:23, Jun-jun Yang wrote:

> I'm a newbie user of pwscf, and working on total energy calculation
> of chromite mineral as my first step for getting familiar with the
> program. [...] The unit cell of the mineral has 56 atoms (Z=8)

you should star with simpler materials if you want to get familiar
with the code 

>    tstress = .true.

don't calculate stress if you do not need to

>     ecutwfc = 40.0

start with a smaller cutoff, increase if needed

>    diagonalization = cg,

davidson is usually faster (but takes more memory)

> K_POINTS automatic
> 8 8 8  0 0 0

start with Gamma only, using all tricks to enhance the speed:

K_POINTS gamma

and increase if needed. The CPU time needed is proportional 
to the number of k-points and to the number of plane waves.

Paolo

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy



More information about the Pw_forum mailing list