[Octopus-users] orthogonalization in td

[Miguel Marques]

  Hi Matthieu

>>  There are no orthogonalizations when you do a TDDFT propagation. The
>> slowdown has a very simple reason: the recalculation of the external
>> potential every time you move the ions (every time-step). This part can
>> be extremely time-consuming (relatively to the time to propagate the
>> orbitals).
> ok - surprising, but ok.

Propagating is a relatively inexpensive N^2 operation. Generating the
external potential is often also implemented as N^2, but can be brought
down by using locality. This, however, involves a lot of bookkeeping,
and has a memory and computational overhead...

>> For the moment, Xavier has optimized already the code, and he
>> has made it parallel in atoms.
> is this turned on by default?

Yes, it is on by default when you are parallel in states (the
parallelization in atoms uses the same communicators as the states). Of
course you can do a mixed states/atoms/domains paralellization.

>> Actually, you should really take a look at the fast
>> Ehrenfest stuff of Xavier et al, that basically allows you to perform a
>> "TDDFT" calculation with a larger time-step.
> ok, this is what I was really interested in - is it present in
> svn-octopus and how do I use it/test it?

Xavier's implementation is basically a one-liner and is controlled by
the variable TDIonicTimeScale

  cheers,
  Miguel

-- 
Dr. Miguel A. L. Marques
marques at tddft.org

Laboratoire de Physique de la Matière Condensée et Nanostructures
(LPMCN) - Université Lyon I
Bâtiment Brillouin, Domaine scientifique de la DOUA
69622 Villeurbanne Cedex
Tel +33 (0)4 72448187
Fax +33 (0)4 72432648