The ABINIT Users Mailing List ( CLOSED )

[Michel Côté <michel.cote@umontreal.ca>]

Title: Re: [abinit-forum] Abinit 5.2.2

Hello Nuno,

Upon looking at the time decomposition that you have provided, it seems that all the time is spent in the timeab routine, which is just the timing routine in Abinit. You may have some problem with the call for time in the timein.F90 subroutine. Each implementation of Fortran seems to have a different to handle the call to time and it is sometimes a problem. It seems to be your case.

Michel


Le 12/09/06 12:05, « Nuno A. G. Bandeira » <nuno.bandeira@ist.utl.pt> a écrit :

> Has anyone had any experience using the latest abinit with gfortran ?
>
> My binaries are painfully slow even with acml.
>
> t01.out in Tests/v1 has this routine result line taking 16 seconds to 
> finish. I'm almost certain this did not happen with the last version of 
> abinit where the timings were competitive with the reference outputs.
>
>
> -----
> - For major independent code sections, cpu and wall times (sec),
> - as well as % of the total time and number of calls
>
> - routine                 cpu     %       wall     %      number of calls
> -                                                          (-1=no count)
> - fourwf(pot)            0.000   NaN      0.159   1.0            320
> - abinit(2)              0.000   NaN      0.156   1.0              1
> - timing timab           0.000   NaN     14.172  87.4             20
> - invars2                0.000   NaN      0.141   0.9              1
> - 62   others            0.000   NaN      0.606   3.7
>
> - subtotal               0.000   NaN     15.234  93.9
>
> ================================================================================
>
>   Calculation completed.
> .Delivered    1 WARNINGs and   1 COMMENTs to log file.
> +Overall time at end (sec) : cpu=          0.0  wall=         16.2


-- 
*************************************************************** 
  Michel Cote                          tel: +1 (514) 343-5628     
  Professeur agrégé                   fax: +1 (514) 343-2071               
  Département de physique 
  Université de Montréal 
  C.P. 6128, succ. Centre-ville 
  Montréal (Québec) H3C 3J7          Michel.Cote@umontreal.ca 
  Canada                   www.phys.umontreal.ca/~michel_cote 
***************************************************************