[Wien] Nd2Fe14B

[Stefaan Cottenier]

> My question now are there is a way to reduce the
> processing time ( it takes about 20 days  to do 20
> iterations as it is for 600 k).

The usual recommendations apply:

* Make sure you have mkl or ATLAS libraries on your pc, to maximize 
processing speed (see the benchmark page 
http://www.wien2k.at/reg_user/benchmark/ to estimate whether you are 
running at optimal speed on your system)

* Choose a model compound that is as simple as possible and 
simultaneously as similar as possible to the compound you want to 
study. Test basis set size and k-mesh on this simple one, to be sure 
that you will do your actual calculations with the smallest basis 
set/k-mesh that give your accurate results.

These are suggestions that are not specific to your problem, but big 
cells need long calculation times, there is no way around that...

Stefaan