[Q-e-developers] [Pw_forum] Increasing the FFT grid spacing/ using wavefunction to provide initial charge density

[Lorenzo Paulatto]

On 30 September 2016 at 11:34, Nicola Marzari <nicola.marzari at epfl.ch>
wrote:

> On 30/09/2016 08:11, Lorenzo Paulatto wrote:
> well, I guess really difficult you mean to make it automatic, correct?
> Automatic wannierization/disentanglement without any human intervention
> is a bit tricky, but in part is also due to the fact that we never put
> a serious effort in it. A few other groups are working on this, including
> Steve Louie in Berkeley, Eric Cances in Paris, and Stefano Curtarolo in
> Duke.
>


Hi Nicola,
I actually mean really difficult from a theory and computation point of
view, even assuming perfect user. It is easy (well, at least it is already
done!) to do Wannier interpolation of the Hamiltonian and other
observables, but it is hard to do it for wavefunction with the constraint
of NOT building explicitly the Wannier functions for the huge supercell
that corresponds to the grid...

Actually, already building explicitly the Wannier function for the initial
grid may require a big supercell and probably won't fit in memory.

And even after you do that, you get a double-sum on k-points that I'm not
sure is any faster than doing SCF for the finer grid directly.

But I may easily be wrong, and there could be non obvious ways to do it
more efficiently.

-- 
Dr. Lorenzo Paulatto
IdR @ IMPMC -- CNRS & Université Paris 6
phone: +33 (0)1 44275 084 / skype: paulatz
www:   http://www-int.impmc.upmc.fr/~paulatto/
mail:  23-24/4é16 Boîte courrier 115, 4 place Jussieu 75252 Paris Cédex 05
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20160930/298c3434/attachment.html>