The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

Hi,

There is likely few tricks here and there that could help a little bit
depending on how well your 32 atoms calculation was set up:
- parallelization
- reduction of the number of k-point since you use a larger cell
- smart SCF algorithms
- nice starting point for the SCF
- tune the CGWF parameters
- ...
Yet Abinit is a order N^3 code. So whatever you do, at best, you'll
get a 512 time longer calculation for the 256 atom cell than for the
32 atom one.

regards

PMA

On Wed, Jun 4, 2008 at 1:32 PM, Toby D. Young <tyoung@ippt.gov.pl> wrote:
>
>
>
> Dear abinit users;
>
> I wish to optimise the structural and electronic components of a large
> super cell (say, 256) atoms. The super cell is to be doped with one
> dopent (hence a large super cell is required).
>
> Can anyone offer some hints, tricks, or suggestions on a reasonable way
> to achieve this aim without utilising large amounts of memory and time?
>
> My experience with a smaller super cell (32 atoms) is already time
> consuming. My approach is currently based on the content of "lesson 3".
>
> I am grateful in advance for responses.
>
> Best,
>        Toby
>
> --
>
> Toby D. Young - Adiunkt (Assistant Professor)
> Department of Computational Science
> Institute of Fundamental Technological Research
> Polish Academy of Sciences
> Room 206, ul. Swietokrzyska 21
> 00-049 Warszawa, Polska
>
> +48 22 826 12 81 ext. 184
> http://rav.ippt.gov.pl/~tyoung
>



-- 
Pierre-Matthieu Anglade