[Pw_forum] Nscf and scf

[Vo, Trinh (388C)]

Dear Stefano,

For scf calculation I see the avg number of iterations varies from 3 to 4
per scf step, and there are total 22 scf steps.  For nscf one, it is about
59.6.  

Thanks,

Trinh 


On 9/6/10 8:05 AM, "Stefano de Gironcoli" <degironc at sissa.it> wrote:

> how many iterations does the iterative diagonalization takes in the nscf
> calculation and how many in the scf  one?
> stefano
>  
> Vo, Trinh (388C) wrote:
>> Dear PWSCF Users,
>> 
>> I run the calculation for a system of 216 atoms (SiP).  The CPU time for the
>> calculation of 4x4x4 kpt mesh, which actually did calculation for 24 kpts due
>> to symmetry, is 1d4h18m (~28h).  When I performed a nscf calculation for
>> 41kpts along 001 direction, the CPU time is 3d5h (~77h).  Thus, the time for
>> one kpt in scf calculation is ~1.17h/kpt,  and 1.87h in nscf calculation.  I
>> am wondering how can it take more time for nscf calculation than for scf
>> calculation.  I don¹t understand why.  Could you explain to me what can cause
>> this to happen.  I use QE-4.1.2.
>> 
>> Thank you very much,
>> 
>> Trinh Vo
>> CalTech/JPL
>> 
>>   
>> ------------------------------------------------------------------------
>> 
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>   
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum