[CP2K-user] [CP2K:21783] Re: SCF converges for outer circle but SCF does not converge in general

[Frederick Stein]

Dear Josh,
You should use the same EPS_SCF for the OUTER_SCF (you did not provide it, 
thus CP2K uses 1.0E-5) as for the inner SCF iteration (in your case 5E-07). 
The iteration did not converged because the inner SCF loop did not converge.
HTH,
Frederick

Joshua Edzards schrieb am Mittwoch, 27. August 2025 um 12:48:34 UTC+2:

> Dear CP2K comunity,
>
> I was running a single-point energy calculation with PBE0 on MOF5 with a 
> hydrogen molecule inside. Somehow, after the third outer SCF loop, the 
> calculation seems to have converged. But the next line clearly states that 
> it failed.
>
>   outer SCF iter =    3 RMS gradient =   0.58E-05 energy =     
>  -1195.0423806694
>   outer SCF loop converged in   3 iterations or   75 steps
>
>
>
>  *******************************************************************************
>  *   ___                                                                   
>     *
>  *  /   \                                                                 
>      *
>  * [ABORT]                                                                 
>     *
>  *  \___/     SCF run NOT converged. To continue the calculation 
> regardless,   *
>  *    |             please set the keyword IGNORE_CONVERGENCE_FAILURE.     
>     *
>  *  O/|                                                                   
>      *
>  * /| |                                                                   
>      *
>  * / \                                                           
>  qs_scf.F:611 *
>
>  *******************************************************************************
>
> I was wondering why this happens. Additionally, I get an error message 
> from Slurm, and I assume that this might cause the problem.
>
> --------------------------------------------------------------------------
> MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
> with errorcode 1.
>
> NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
> You may or may not see output from other processes, depending on
> exactly when Open MPI kills them.
> --------------------------------------------------------------------------
> [c0365:144409] 191 more processes have sent help message help-mpi-api.txt 
> / mpi-abort
> [c0365:144409] Set MCA parameter "orte_base_help_aggregate" to 0 to see 
> all help / error messages
>
> The Slurm script, the error, and also CP2K in- and output files are 
> attached. Any help is appreciated. Also, if more information is needed, I 
> am happy to provide it.
>
> Thank you very much, and all the best
> Josh
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/2793a9e5-7148-4f86-a086-825e79eb3c26n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250827/a79a0f4e/attachment-0001.htm>