[QE-users] problem in cp.x parallel calculation version 6.4 and 6.3
Zahra Nourbakhsh
z.nourbakhsh at gmail.com
Thu May 7 12:01:12 CEST 2020
Dear Lorenzo
Yes, I know many people use cp.x parallel but as I mentioned this problem
occurs for me when I use version 6.4 and 6.3; for qe-6.1, I don't have this
problem and I don't know about other versions.
Here are some lines of my BOMD calculations using
srun -n 16 cp.x <in
and
srun -n 32 cp.x <in
for 16 cores:
nfi ekinc temph tempp etot enthal
econs econt vnhh xnhh0 vnhp
xnhp0
nfi tempp E -T.S-mu.nbsp +K_p #Iter
Step 1 0 -550.100617 -550.100617 -550.100617 42
Step 2 0 -550.100617 -550.100617 -550.100617 6
Step 3 0 -550.100617 -550.100617 -550.100617 6
MAIN: EKINC (thr) DETOT (thr) MAXFORCE
(thr)
MAIN: 0.000000D+00 0.1D-05 0.484306D-10 0.1D-07 0.000000D+00
0.1D+11
MAIN: convergence achieved for system relaxation
* Physical Quantities at step: 4
Pressure of Nuclei (GPa) 0.00000 4
Pressure Total (GPa) -3.80754 4
NOTE: eigenvalues are not computed without ortho
total energy = -550.10061707572 Hartree a.u.
kinetic energy = 405.35380 Hartree a.u.
electrostatic energy = -595.48960 Hartree a.u.
esr = 1.06694 Hartree a.u.
eself = 625.54150 Hartree a.u.
pseudopotential energy = -286.39030 Hartree a.u.
n-l pseudopotential energy = 61.79974 Hartree a.u.
exchange-correlation energy = -135.37426 Hartree a.u.
average potential = 0.00000 Hartree a.u.
.
.
.
----------------------------------------------------------------------------------------------------
for 32 core:
nfi ekinc temph tempp etot enthal
econs econt vnhh xnhh0 vnhp
xnhp0
nfi tempp E -T.S-mu.nbsp +K_p #Iter
Step 1 0 -523.513474 -523.513474 -523.513474 42
Step 2 0 -525.555463 -525.555463 -525.555463 21
Step 3 0 -527.873907 -527.873907 -527.873907 20
Step 4 0 -529.108725 -529.108725 -529.108725 20
Step 5 0 -529.707861 -529.707861 -529.707861 20
Step 6 0 -530.035707 -530.035707 -530.035707 20
Step 7 0 -530.220064 -530.220064 -530.220064 20
and it continues up to Step 50 while I have run cp.x for BO system where
ion_velocity=0 and electron_dynamics = 'cg'.
and if I run the program on higher cores, for example, srun -n 64 cp.x
nfi tempp E -T.S-mu.nbsp +K_p #Iter
Step 1 0 -439.090690 -439.090690 -439.090690 46
Step 2 0 NaN NaN NaN 100
Step 3 0 NaN NaN NaN 100
*my input file is as follow*
&control
title = ' Water 32 molecules ',
calculation = 'cp',
restart_mode = 'from_scratch', ! 'restart',
ndr = 50,
ndw = 50,
! nstep = 100,
! iprint = 10,
! isave = 100,
tstress = .TRUE.,
tprnfor = .TRUE.,
! dt = 3.0d0,
etot_conv_thr = 1.d-8,
prefix = 'h2o',
pseudo_dir='/u/nzahra/pseudo',
outdir='/ptmp/nzahra/water/bo/scan/32/blyp',
/
&system
ibrav = 14,
celldm(1) = 18.65,
celldm(2) = 1.0,
celldm(3) = 1.0,
celldm(4) = 0.0,
celldm(5) = 0.0,
celldm(6) = 0.0,
nat = 96,
ntyp = 2,
nbnd = 128,
ecutwfc = 70.0,
input_dft='BLYP'
/
&electrons
electron_dynamics = 'cg',
! electron_velocities = 'zero',
/
&ions
ion_dynamics = 'none',
ion_velocities = 'zero',
ion_radius(1) = 0.8d0,
ion_radius(2) = 0.5d0,
/
&cell
cell_dynamics = 'none',
/
ATOMIC_SPECIES
O 16.0d0 O.blyp-mt.UPF
H 1.0079d0 H.blyp-vbc.UPF
ATOMIC_POSITIONS (crystal)
O 0.378966142 0.399351787 0.153232022
O 0.674080975 0.071652846 0.096431003
O 0.502725138 0.249767593 0.960543532
O 0.870124149 0.014393000 0.900181919
O 0.370188926 0.134633533 0.543797631
O 0.890478795 0.251227789 0.734444338
O 0.942635278 0.708667766 0.077711606
O 0.525930248 0.628961766 0.190669923
O 0.824014650 0.270393021 0.220728713
O 0.759518359 0.515784572 0.070294996
O 0.623729416 0.669186389 0.445331655
O 0.680125097 0.418204412 0.825978850
O 0.996082615 0.864590486 0.308543405
O 0.770361333 0.914797185 0.470010581
O 0.945419200 0.594216294 0.657975197
O 0.067635795 0.453586124 0.286217145
O -0.035358869 0.169241966 0.480014817
O 0.779742615 0.439946163 0.474919680
O 0.645018862 0.163367314 0.499583173
O 0.305792605 0.499856745 0.608241857
O 0.130443525 0.393912011 0.788840284
O 0.229230973 0.237679258 0.336167096
O 0.285232159 0.176280147 0.808692313
O 0.475886585 0.903937356 0.211739054
O 0.173759156 0.531282254 0.024620567
O 0.209963773 0.999982487 0.194716580
O 0.121031392 0.668269158 0.455966686
O 0.284992985 0.749172203 0.852530295
O 0.539738019 0.637507450 0.704744940
O 0.363201790 0.862178393 0.486282852
O 0.778560789 0.808717073 0.723047518
O 0.145671119 0.973743672 0.921717439
H 0.426295336 0.340207686 0.086508715
H 0.341110554 0.340071056 0.223999914
H 0.901742783 0.187895961 0.406332781
H 0.614247468 0.002591432 0.137160396
H 0.737402864 0.029853334 0.031561454
H 0.562667835 0.181611878 1.003775164
H 0.564472433 0.307913217 0.906055245
H 0.870911680 0.095248333 0.841283158
H 0.361495596 0.035370667 0.530609224
H 0.341028286 0.151696206 0.638151234
H 0.722481104 0.002887488 0.473461397
H 0.905090080 0.219954239 0.640293326
H 0.954076861 0.766015557 0.158801634
H 0.610163252 0.596939823 0.147107980
H 0.466232471 0.548685433 0.185384792
H 0.765821036 0.195960162 0.185901401
H 0.912890221 0.253846698 0.182869659
H 0.785503768 0.438364362 0.126964918
H 0.831590067 0.586920501 0.079653737
H 0.685923292 0.590588094 0.448446869
H 0.574485302 0.655736601 0.358925659
H 0.750458474 0.361169500 0.783295312
H 0.720257953 0.455946686 0.910136383
H 1.038756633 0.795539302 0.367591617
H 0.919030853 0.899810647 0.361832880
H 0.836130766 0.940160626 0.842150509
H 0.705435215 0.844189019 0.443076318
H 0.888087349 0.668069262 0.695421637
H 0.126921868 0.376078035 0.307926241
H 0.088965741 0.478032540 0.192394629
H 0.053662315 0.190925279 0.443309179
H 0.844720849 0.472080648 0.542930479
H 0.833371218 0.414408856 0.396199674
H 0.547079627 0.164405051 0.517471345
H 0.676341327 0.256807216 0.498799966
H 0.261012491 0.567734272 0.551837812
H 0.385073439 0.545908944 0.646744137
H 0.190247229 0.440084410 0.720831264
H 0.183107423 0.307801224 0.803750912
H -0.026090029 0.302102511 0.754164090
H 0.282273377 0.209449298 0.418128934
H 0.222760066 0.154169627 0.279980690
H 0.235413166 0.097779415 0.848240164
H 0.365798276 0.193666507 0.865859469
H 0.381805142 0.932319112 0.194456804
H 0.482699112 0.808034933 0.185394178
H 0.030863614 0.666924190 0.062015985
H 0.252956198 0.485820135 0.067028091
H 0.185028596 0.993855524 0.097525880
H 0.134140048 0.954300465 0.242452683
H 0.060467045 0.649466748 0.533895806
H 0.104799632 0.590923687 0.392958285
H 0.263687068 0.672221596 0.911020950
H 0.369208033 0.727116113 0.806274112
H 0.591365151 0.557840346 0.736279678
H 0.560587315 0.649238296 0.606795414
H 0.153193921 0.477570557 0.941851144
H 0.281022580 0.809008014 0.474911677
H 0.406393144 0.864434913 0.396869127
H 0.688858471 0.767344468 0.733825538
H 0.776549080 0.853385230 0.632459869
H -0.012423931 0.546193409 0.732603372
H 0.185208669 0.887687972 0.889096725
H 0.046119603 0.973148479 0.910544721
Yours sincerely,
Zahra Nourbakhsh
On Thu, May 7, 2020 at 1:23 AM Zahra Nourbakhsh <z.nourbakhsh at gmail.com>
wrote:
> Dear experts,
>
> I have just started BOMD calculations using the QE package.
>
> I found that for QE version 6.3.0 and 6.4.1, the cp.x command does not
> work properly when I run it parallel and increase/decrease the number of
> cores. I have run the same input from scratch, but the results are core
> dependent!
>
> But this issue does not exist for version 6.1.0 and using this version, my
> results are the same for any numbers of cores.
>
> I have asked our cluster supporting team about this issue but they believe
> that it comes from QE. Does anyone have any suggestions?
>
> Yours sincerely,
> Zara
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20200507/f402d45e/attachment.html>
More information about the users
mailing list