[Pw_forum] phonon calculations in a spercell of 100 atoms

Paolo Giannozzi giannozz at democritos.it
Mon Feb 13 21:45:09 CET 2012


On Feb 13, 2012, at 16:41 , Lorenzo Paulatto wrote:

> If you only need the phonons at  the gamma point, it may be faster
> to get them by finite displacement, doing 3*nat pw.x ground state  
> calculations.

faster? not sure, and not by much anyway. A phonon calculation at q=0  
also
requires 3*nat self-consistent linear-response calculations. Each  
such calculation
has a computational cost comparable to that of a self-consistency.  
Usually one
needs to perform two finite-difference calculations with opposite  
displacements
in order to get good numerical accuracy, by the way. Or at least,  
this is the way
I would do it.

For q=0 phonons, using k=0 only for the Brillouiin zone sampling, and  
norm-conserving
PP, there is the simplified version of the phonon code that is  
definitely faster and takes
(relatively) little memory

P.
---
Paolo Giannozzi, Dept of Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222







More information about the users mailing list