[Q-e-developers] Fwd: Re: [Q-e-commits] r10603 - trunk/espresso/PW/src

[Lorenzo Paulatto]

On 11/18/2013 06:44 PM, giannozz wrote:
> Author: giannozz
> Date: 2013-11-18 18:44:11 +0100 (Mon, 18 Nov 2013)
> New Revision: 10603
> 
> Modified:
>    trunk/espresso/PW/src/electrons.f90
> Log:
> More exx cleanup: the self-consistency loop at fixed exchange potential is no 
> longer intertwined with the loop over exchange. Calculation and printout of
> the various energy terms is less confusing (in my opinion). There is still 
> a small but non-negligible difference between gamma-only and k=0, but this
> should not be related to these changes.
> 
> 

About this subject: it is smart to have two concentric SCF cycles, one
to converge the wavefunctions and one that exppensively updates the Fock
operator when they are converged.

However, if you set up the equally smart adaptive threshold the Fock
operator will still be updated at almost every cycle.

So now we have two smart tricks, each of them extremely complicated, but
I feel that they almost exactly cancel each other out.

Anybody with more experience on real-world EXX calculations cares to
comment?

cheers


-- 
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 5