[Columbus] Error in MCSCF optimization

[Hans]

Dear Filipe,

There is not enough memory. You need 1.4 billion more 8*byte words more. 
Please check also your CI dimension in cidrtls.

Best regards, Hans

On 3/20/2017 1:48 AM, Filipe Belarmino wrote:
> Dear developers for Columbus7.0.,
>
>
> I have tried to do a MCSCF optimization with CI gradients for 4 states.
> The energies in mcscfsm.sp are ok, but after that the calculation
> doesn't continue. I have got the following error:
>
>
> =====================================================================
> using                      1  npartitions of size                     1 .
>
> pliste=   0
>  using                      1  npartitions of size                     1 .
> pliste=   0
>  Main memory management:
>  global                1 DP per process
>  vdisk                 0 DP per process
>  stack                 0 DP per process
>  core         4194303999 DP per process
>  bummer (warning):resetting maxbl4 to 2**20-1                1048575
>
>  intermediate da file sorting parameters:
>  nbuk=   2 lendar=   32767 nipbk=   21844 nipsg=4190363142
>   waiting after driverwrap ....                     0
>  Allocating             4194303999  for ci step ...
>  bummer (fatal):strefv: (loc3-totfre)=             1384359863
> bummer
> ===================================
> #################################################
>
>
> The ciudgin was thus made:
>
>  &input
>  NTYPE = 0,
>  GSET = 0,
>   DAVCOR =10,
>  NCOREL = 58
>  NROOT = 1
>  IVMODE = 3
>  NBKITR = 1
>  NVBKMN = 1
>  RTOLBK = 1e-4,
>  NITER = 30
>  NVCIMN = 1
>  RTOLCI = 1e-4,
>  NVCIMX = 3
>  NVRFMX = 3
>  NVBKMX = 3
>  IDEN  = 1
>  CSFPRN = 10,
> /&end
>
> ####################################
>
>
> How can I fix that?
>
>
> Sincerely,
>
>
> Filipe Belarmino.
>
> Departamento de Química
> Universidade Federal da Paraíba
>
>
>
>
> _______________________________________________
> Columbus mailing list
> Columbus at lists.osc.edu
> https://lists.osu.edu/mailman/listinfo/columbus
>