[Wien] Large no of K-points in optical calculations

[Peter Blaha]

a) It seems you are still using WIEN97 ?

b) When you use up to 10,000 k-points, try to increase nkpt exactly up to
this value (and not to the larger 15,000)

c) Aparently with 15,000 the program is larger than your memory or size-limit
of your computer.  You must adjust the parameters (and the number of k-points
in the IBZ!!!!) to the limits of your computer.


> I am trying to calculate optical properties with up to 10,000 k-points in the IBZ. However, while executing
> x joint -up I got the following : STOP  nst. gt. nkpt.
>
> I then had to increase the no of k-points in SRC_joint/param.inc to 15,000 and the following errors were obtained when I executed 'make'  in arbdos.f:
>
> Line 23: REAL *4  OPMAT
> Array "opmat" at (^) is too large to handle.
> arbdos.f : outside of any programs unit.
> arbdos: 23:size of variable "opme_" is too large
>
> I went further to reduce the no  NKPT in SRC_joint/param.inc to 5000 but now got segmentation fault when I excuted  x joint -up
>
> Please I need your suggestion on how to get over this.
>
>  Emeka
>
>
>
>
> Chukwuemeka M. I. Okoye,
>
> Department of Physics and Astronomy,
>
> University of Nigeria,Nsukka,
>
> Enugu State,NIGERIA
>
> 234-042-771611
>
>
> ---------------------------------
> Do you Yahoo!?
> New Yahoo! Photos - easier uploading and sharing


                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------