forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] scaling

From: "Anglade Pierre-Matthieu" <anglade@gmail.com>
To: forum@abinit.org
Subject: Re: [abinit-forum] scaling
Date: Mon, 13 Nov 2006 17:01:32 +0100
Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=m2FMdvZhUj3vF5Kvp50Qst7xlDVU7cSCuczBC/LOzUTHtyQGWyYs6rFS/6J48O+9i9exOxTp5sFIthcrVIcgxEzUG5qcbAsp94q6P5cizHvpHHz3aO/MY8XRR69e8iX0QaSMmFYKS8f+5T6OfrKs4MftsyK5PsS0Nt3nfzRmJTc=

Hi,
the memory requirement depends mainly on:
-nband : proportional
-acell : cubic
- nkpt : proportional
-ecut: cubic
-rprim (because it determines the cell volume when combined with acell)
-many other parameters like for instance mixing algorithm.
PMA

On 11/13/06, Stephan Kremers <kremers@physik.rwth-aachen.de> wrote:

Hello,

I know that the memory, wich is needed for the calculations, scales with
the number of valence electrons.
Also I have seen, that the memory rises with bigger sizes of the cell
(acell)

Does anyone know, which other parameters are responsible for the
required memory? And how is the relationship between the required memory
and these parameters?

greetings
Stephan Kremers

--
Pierre-Matthieu Anglade

scaling, Stephan Kremers, 11/13/2006
- Re: [abinit-forum] scaling, Anglade Pierre-Matthieu, 11/13/2006
- Re: [abinit-forum] scaling, Xavier Gonze, 11/17/2006