large water system calculation with cp2k

Phil G. phil... at
Mon Sep 26 09:19:39 CEST 2016

Dear Samuel,

two further attempts with curvy steps (the TRS4 method was turn off) with 
and without S-PRECONDITIONER ATOMIC and both cases lead to the error:

 *** 15:15:55 ERRORL2 in dm_ls_scf:ls_scf_init_read_write_input processor 
 *** 0  :: err=-300 S Preconditioning not implemented in combination with 
 *** CURVY_STEPS. dm_ls_scf.F line 738                                    
What is the reason for this error and what the code should look in the 
section LS_SCF ?
In one of the last posts there is an input file and you can see what is 
missing and what is wrong in the code.

Kind regards,


Am Mittwoch, 21. September 2016 15:31:01 UTC+2 schrieb Samuel Andermatt:
> From your calculation it seems you did TRS4 and not curvy steps. It seems 
> that the convergence diverged. A solution to this could be to make the 
> mixing fraction smaller or use curvy steps. By the way, if you do not want 
> to waste too much computer time. Already at step 6 of the first iteration 
> it was clear that the run will not converge. At this point you could cancel 
> the calculation. Generally stop the calculation if the energy becomes 
> positive.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the CP2K-user mailing list