[CP2K-user] Problem when calculate slab model, the memory usage is increased and finally it stops

[Бранислав Миловановић]

Dear Ke, Developers,

I have the same problem with increasing memory usage over time, with the 
following input used:

&GLOBAL
  PROJECT Na+_G_quad_wannier
  RUN_TYPE MD
  PRINT_LEVEL LOW
  PREFERRED_FFT_LIBRARY FFTSG
&END GLOBAL

&FORCE_EVAL
  METHOD QS
  &DFT
    BASIS_SET_FILE_NAME 
/home/branislavm/Projects/Programs/Potentials/BASIS_SET
    POTENTIAL_FILE_NAME 
/home/branislavm/Projects/Programs/Potentials/GTH_POTENTIALS
    CHARGE 1
    &MGRID
      CUTOFF 320
      REL_CUTOFF 50
      NGRIDS 4
    &END MGRID
    &QS
      METHOD GPW
      EPS_DEFAULT 1.0E-12
    &END QS
    &SCF
      MAX_SCF 100
      EPS_SCF 5.0E-7
      SCF_GUESS RESTART
      &OT ON
        MINIMIZER DIIS
        N_DIIS 7
        PRECONDITIONER FULL_ALL
        ENERGY_GAP 0.001
      &END OT
    &END SCF
    &XC
      &XC_FUNCTIONAL BLYP
      &END XC_FUNCTIONAL
      &VDW_POTENTIAL
        DISPERSION_FUNCTIONAL PAIR_POTENTIAL
        &PAIR_POTENTIAL
          TYPE DFTD3
          PARAMETER_FILE_NAME 
/home/branislavm/Projects/Programs/Potentials/dftd3.dat
          REFERENCE_FUNCTIONAL BLYP
        &END PAIR_POTENTIAL
      &END VDW_POTENTIAL
      &XC_GRID
        XC_SMOOTH_RHO NN10
        XC_DERIV SPLINE2_SMOOTH
      &END XC_GRID
    &END XC
    &LOCALIZE
      METHOD JACOBI
      MAX_ITER 10000
      &PRINT
        &WANNIER_CENTERS
          IONS+CENTERS
          FILENAME =WANNIER_CENTERS.xyz
          &EACH
            MD 1
          &END
        &END WANNIER_CENTERS
      &END PRINT
    &END LOCALIZE
  &END DFT
  &SUBSYS
    &CELL
      ABC 25.0 25.0 25.0
      PERIODIC NONE
    &END CELL
    &COORD
  N        10.4441985755        7.2694831234       12.5073686365
  C        11.7536307003        7.7158627265       12.7113477991
  C        12.7183618726        6.7163785851       12.3377997831
*   .**                     .**                              .**            
                  .*
*   .**                     .**                              .**            
                  .*
*   .                     .**                              .**              
                .*
 Na        12.4370558895       12.2824472182       12.9521881866
    &END COORD
    &KIND H
      BASIS_SET DZVP-GTH-BLYP
      POTENTIAL GTH-BLYP-q1
    &END KIND
    &KIND O
      BASIS_SET DZVP-GTH-BLYP
      POTENTIAL GTH-BLYP-q6
    &END KIND
     &KIND C
      BASIS_SET DZVP-GTH-BLYP
      POTENTIAL GTH-BLYP-q4
    &END KIND
    &KIND N
      BASIS_SET DZVP-GTH-BLYP
      POTENTIAL GTH-BLYP-q5
    &END KIND
    &KIND Na
      BASIS_SET DZVP-GTH-BLYP
      POTENTIAL GTH-BLYP-q9
    &END KIND
  &END SUBSYS
&END FORCE_EVAL

&MOTION
  &MD
    ENSEMBLE REFTRAJ
    STEPS 20000
    &REFTRAJ
      EVAL_ENERGY_FORCES
      TRAJ_FILE_NAME REFTRAJ.xyz
    &END REFTRAJ
  &END MD
  &PRINT
    &RESTART
      ADD_LAST NUMERIC
      &EACH
        MD 500
      &END
    &END RESTART
  &END PRINT
&END MOTION

I run:

CP2K version 5.1
SVN source code revision svn:18091
cp2kflags: libint fftw3 libxc parallel mpi3 scalapack libderiv_max_am1=5 
libint_max_am=6

on the machine with:

AMD Ryzen Threadripper 1950X 16-Core Processor
G.skill DDR4 2x16GB 3200MHz RipJaws V CL15, F4-3200C15D-32GVK

When I used older cp2k version (2.5.1) I never had similar problems.

Best,
Branislav

среда, 30. октобар 2019. 10.08.34 UTC+1, Ke Zhou је написао/ла:
>
> Dear Developers,
>
> I have ask the questions of increasing memory usage for several times. Now 
> I can calculate bulk system without stoping. 
>
> But now I found the *memory usage for cp2k is increased with time (*for both 
> of version 5.1 and 6.1*)*during running when I calculate the *slab-model 
> system (water in 2D channel )* . Finally,  it stops when the usage reach 
> the upper bound. I also tried cp2k 5.1 and had same problem. *But I run 
> it in our old cluster which is 2.5.1 version, it works well.*
>
> However, I run one of the example in CP2K package, H2O-64.inp (path: 
> ./test/QS/benchmark), which is bulk water system, using both 6.1 and 5.1, 
> it runs well.
>
> Can you help me?
> Is the problem from compile (may be some libs) or the input file?
>
> ########## my input file,  the memory usage for cp2k is increased with time
>  ##################
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191130/7bd0607d/attachment.htm>