[Pw_forum] problems calculating large numbers of empty states

[Nicola Marzari]

Paolo Giannozzi wrote:
> On Oct 22, 2009, at 21:29 , Brad Malone wrote:
> 
>> Hi, I'm trying to calculate a large number of conduction band
>> states for cubic Si (to be used in a subsequent GW calculation).
>> However, when doing this I sometimes run into errors like
>> "problems computing cholesky decomposition"  or "too many
>> bands are unconverged".
> 
> the algorithm used in PWscf is based on the assumption that
> the number of Kohn-Sham states (aka bands) is much smaller
> than the number of basis functions (i.e. plane waves). If this
> assumption doesn't hold (and it doesn't if you calculate 500
> bands in fcc Si unless you use a very large cutoff) you may
> run into trouble. Moreover iterative diagonalization becomes
> slower and more memory-consuming than conventional
> diagonalization. If you really need to do that, you need to
> modify the code


Hi Paolo, Brad,


could one use CP instead - i.e. are the files cross-readable
between the codes ?

If they were, could one do a calculation with CP and 4 valence bands,
all full, and then restart that calculations requesting 500 bands,
frozen Hamiltonian, and all bands filled. Then, a damped dynamics or
conjugate gradients should converge all those 500 filled bands, in
the frozen/correct Hamiltonian, to the ground state ?

				nicola


-- 
---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu