[CP2K:8075] SCF, MD run-time verses atomic species.

hut... at chem.uzh.ch hut... at chem.uzh.ch
Thu Aug 18 14:05:53 CEST 2016


Hi

The PBE-D3 setup looks fine.

Yes, 10-20 SCF iterations during an MD are on the very high end for
such a well behaved system. I would try the following setup (in case you
don't use it already)

&QS
  ....
  EXTRAPOLATION ASPC
  EXTRAPOLATION_ORDER  4
&END QS

&OT
   PRECONDITIONER FULL_SINGLE_INVERSE
   MINIMIZER DIIS
&END OT

regards

Juerg
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut... at chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp... at googlegroups.com wrote: -----To: cp2k <cp... at googlegroups.com>
From: Simiam Ghan 
Sent by: cp... at googlegroups.com
Date: 08/18/2016 01:08AM
Subject: Re: [CP2K:8075] SCF, MD run-time verses atomic species.

Hello Juerg,Thank you for your reply.  I added REFERENCE_C9_TERM .TRUE.  to the vdw section as you suggested and found that my MD simulation became twice (!) as fast.   That is, time per MD step went from ~60 sec to ~30 sec for box of water with 64 molecules and K,Cl ions.  I made a quick comparison of Energy_Force results with and without this flag and indeed forces and energies do not change significantly.   This is great news.  Why is this not a default?  What's the catch?   Also, would you say this is now a correct way to declare a PBE-D3 setup?

&VDW_POTENTIAL         POTENTIAL_TYPE  PAIR_POTENTIAL         &PAIR_POTENTIAL           R_CUTOFF     1.5000000000000005E+01           TYPE  DFTD3           PARAMETER_FILE_NAME dftd3.dat           REFERENCE_FUNCTIONAL PBE           CALCULATE_C9_TERM  T           REFERENCE_C9_TERM  T         &END PAIR_POTENTIAL       &END VDW_POTENTIAL
Finally, I mentioned earlier that 30 scf steps were needed to converge the same box of water with ions.   I was then using EPS_SCF = 1E-06.   Reducing this to EPS_SCF = 1E-05 brought the number down to 10-20 scf necessary.  Does that still sound high?   Is the definition of EPS_SCF documented somewhere, as it is apparently not just Energy convergence. 
Cheers,Simiam Ghan



On Friday, July 1, 2016 at 3:10:52 PM UTC+3, jgh wrote:Hi



no this shouldn't be, but without more information I will have to guess.

You could also have a look at the timings at the end of the output to 

see if some routines got slower or if all parts of the run were affected.



Two things to consider:



1) Use REFERENCE_C9_TERM TRUE in order to reduce the time for vdW in MD.

2) 30 SCF iterations in MD for such a simple system is pointing to a problem

   with your setup. 



regards



Juerg

--------------------------------------------------------------

Juerg Hutter                         Phone : ++41 44 635 4491

Institut für Chemie C                FAX   : ++41 44 635 6838

Universität Zürich                   E-mail: hut... at chem.uzh.ch

Winterthurerstrasse 190

CH-8057 Zürich, Switzerland

---------------------------------------------------------------



-----cp... at googlegroups.com wrote: -----To: cp2k <cp... at googlegroups.com>

From: Simiam Ghan 

Sent by: cp... at googlegroups.com

Date: 07/01/2016 01:38PM

Subject: [CP2K:7883] SCF, MD run-time verses atomic species.



Dear all,I am running NVT MD with Quickstep on a box of water with 64 molecules.   If I replace a water molecule with a KCl ion pair, i observe that the MD and SCF step times more than double on my setup.  (MD from ~20 sec to ~46 sec). SCF iterations are converging (except the very first MD step) in around 30 steps in both cases but each SCF now takes over twice as long as before.   Is there an explanation of why such a 'small' change in system could double the run time?  Is KCl really so heavy to calculate compared to H2O?  The number of electrons in the system increases by 8, from 512 to 520.  Below my KCl input file. 

Greetings,Simiam





&GLOBAL

  PROJECT H2O_KCl  RUN_TYPE MD  PRINT_LEVEL MEDIUM

&END GLOBAL

&FORCE_EVAL

  METHOD Quickstep                                ! GPW method.

  &SUBSYS                                       ! A subsystem: coordinates, topology, molecules and cell.

    &CELL                                       ! Supercell setup.      ABC [angstrom] 12.414 12.414 12.414        ! Using 64 H2O molecules, we thus get a density of 1g/cm^3.      PERIODIC XYZ                                  ! Use PBC in all dimensions.    &END CELL

    &COORD    UNIT angstromH -0.567712 -0.469646 -0.645913H 0.626116 -0.687796 0.308193O 0 0 0(...)K 2.1035 2.1035 4.2735Cl 4.1035 4.1035 2.6035                 ###H 3.73881 3.10388 5.46104

H 3.65742 2.89924 6.989O 3.1035 3.1035 6.207(...)    &END COORD

    &KIND O      BASIS_SET DZVP-MOLOPT-GTH-q6      POTENTIAL GTH-PBE-q6    &END KIND    &KIND H      BASIS_SET DZVP-MOLOPT-GTH-q1      POTENTIAL GTH-PBE-q1    &END KIND     &KIND K      BASIS_SET DZVP-MOLOPT-SR-GTH-q9      POTENTIAL GTH-PBE-q9    &END KIND    &KIND Cl      BASIS_SET DZVP-MOLOPT-GTH-q7      POTENTIAL GTH-PBE-q7    &END KIND









  &END SUBSYS

 &DFT

    BASIS_SET_FILE_NAME  BASIS_MOLOPT    POTENTIAL_FILE_NAME  GTH_POTENTIALS    !    SPIN_POLARIZED                ! Do spin-polarized calculation

    &POISSON       PERIODIC XYZ    &END POISSON

    &QS      METHOD GPW      EPS_DEFAULT 1.0E-10   ! Set various epsilons for QS to values that will lead                                    ! to energy correct up to 1e-10.    &END QS

    &MGRID      CUTOFF 400    ! This is Ecut of eq. 39 in VandeVondele (2005), i.e., plane-wave cutoff                               ! that determines size of finest grid (see caption of Fig. 1). Cutoffs for                    ! the subsequent, coarser grid levels are given by eq. 39.      NGRIDS 4      ! This is N of eq. 39 in VandeVondele (2005), i.e., number of grids used.      REL_CUTOFF 40 ! This controls the grid level onto which Gaussians will be mapped.    &END MGRID

    &XC

      &XC_FUNCTIONAL        &PBE                           PARAMETRIZATION ORIG        &END PBE      &END XC_FUNCTIONAL

      &VDW_POTENTIAL

         POTENTIAL_TYPE PAIR_POTENTIAL

         &PAIR_POTENTIAL            TYPE DFTD3            REFERENCE_FUNCTIONAL PBE            CALCULATE_C9_TERM .TRUE.            PARAMETER_FILE_NAME dftd3.dat            R_CUTOFF 15.0         &END PAIR_POTENTIAL

      &END VDW_POTENTIAL

    &END XC

    &SCF

      SCF_GUESS RESTART                ! Use data from previous run as initial guess for wavefunction.      EPS_SCF 1.0E-6                ! Threshold for converged total energy.      MAX_SCF 300                ! Maximum number of SCF iterations performed.

      &OT        PRECONDITIONER NONE        ! This should be stable with respect to the "Cholesky errors"      &END OT

       &PRINT         &RESTART ON

                BACKUP_COPIES 1

                       &EACH                        MD 1                       &END EACH

                ADD_LAST NUMERIC

         &END RESTART

       &END PRINT

    &END SCF

    &PRINT      &E_DENSITY_CUBE

        STRIDE 1 1 1

        &EACH                    MD 99999999        &END EACH

        ADD_LAST NUMERIC     &END E_DENSITY_CUBE

      &PDOS            COMPONENTS .FALSE.            NLUMO = -1            FILENAME dosfile            LOG_PRINT_KEY TRUE

            &EACH                MD 99999999            &END EACH

            ADD_LAST NUMERIC      &END PDOS

    &END PRINT

 &END DFT

&END FORCE_EVAL

&MOTION

        &MD                ENSEMBLE        NVT                STEPS                10000                TEMPERATURE        300.0        ! K                TIMESTEP        0.5        ! fs                                &THERMOSTAT

                        REGION GLOBAL                        TYPE NOSE

                        &NOSE                                LENGTH 3        ! Length of Nose-Hoover chain                                TIMECON 20.0        ! Period of typical vibrational motion in system in fs                        &END NOSE                                                                &END THERMOSTAT                        &END MD

        &PRINT

                &RESTART

                        &EACH                              MD 1                        &END EACH

                        ADD_LAST NUMERIC

                &END RESTART

                &TRAJECTORY ON                           ADD_LAST NUMERIC                           FILENAME trajectory                &END TRAJECTORY

        &END PRINT

&END MOTION











-- 



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+... at googlegroups.com.



To post to this group, send email to cp... at googlegroups.com.



Visit this group at https://groups.google.com/group/cp2k.



For more options, visit https://groups.google.com/d/optout.









-- 

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+uns... at googlegroups.com.

To post to this group, send email to cp... at googlegroups.com.

Visit this group at https://groups.google.com/group/cp2k.

For more options, visit https://groups.google.com/d/optout.





More information about the CP2K-user mailing list