This thread is locked.Only browsing is available.
Top Page > Browsing
problems with MD in OpenMX3.2
Date: 2007/04/10 14:21
Name: Denis Music   <music@mch.rwth-aachen.de>

Dear Prof. Ozaki and OpenMX users,
I've been running MD (NVT_VS) with OpenMX3.1 and had no problems. With the latest version, the temperature from 3000 K (set point) increased to extreme values during the run:
time= 0.000 (fs) Energy= -224.47455 (Hatree) Temperature= 0.000
time= 1.000 (fs) Energy= -125.06451 (Hatree) Temperature= 2973.937
time= 2.000 (fs) Energy= -106.08964 (Hatree) Temperature= 40365264.595
time= 3.000 (fs) Energy= -293.04287 (Hatree) Temperature= 263829964.790
time= 4.000 (fs) Energy= -335.26899 (Hatree) Temperature= 616964567.914
Something seems to be wrong in OpenMX3.2. Could you please help me out?
With best regards,
Denis
メンテ
Page: [1]

Re: problems with MD in OpenMX3.2 ( No.1 )
Date: 2007/04/10 20:21
Name: T.Ozaki

Hi,

If you do not mind, could you show us your input file ?

Regards,

TO
メンテ
Re: problems with MD in OpenMX3.2 ( No.2 )
Date: 2007/04/11 15:00
Name: Denis Music  <music@mch.rwth-aachen.de>

Hello again,
Of course I don't mind. Here is the input file.

#
# File Name
#
System.CurrrentDirectory ./ # default=./
System.Name out
level.of.stdout 1 # default=1 (1-3)
level.of.fileout 1 # default=1 (1-3)
#
# Definition of Atomic Species
#
Species.Number 1
<Definition.of.Atomic.Species
B B6.5-s1p1 B_PBE
Definition.of.Atomic.Species>
#
# Atoms
#
Atoms.Number 96
atoms.SpeciesAndCoordinates.Unit Ang
<Atoms.SpeciesAndCoordinates
1 B 2.3428329 -2.0529018 -3.8011353 1.50000 1.50000
2 B -0.7735232 1.5263710 3.0210434 1.50000 1.50000
3 B 4.1080964 -0.3495180 -0.8061285 1.50000 1.50000
4 B 3.9972693 0.6996831 3.3965726 1.50000 1.50000
5 B 3.0159726 -2.4432895 1.6699144 1.50000 1.50000
6 B 3.1000425 2.8670685 4.4694455 1.50000 1.50000
7 B 0.9658347 1.5304652 2.5149826 1.50000 1.50000
8 B 4.0266769 1.8838434 1.9082262 1.50000 1.50000
9 B 1.3690943 1.5441100 0.1049908 1.50000 1.50000
10 B 0.1190549 1.9223794 6.4038808 1.50000 1.50000
11 B 3.2224223 3.3739232 2.5232715 1.50000 1.50000
12 B 1.3528642 2.6890132 4.3431837 1.50000 1.50000
13 B 5.2063076 5.8974602 -0.8489882 1.50000 1.50000
14 B 3.5508526 -2.7882185 5.0780803 1.50000 1.50000
15 B 8.0267453 0.0686714 0.4975720 1.50000 1.50000
16 B -1.3777318 0.3864925 4.3361944 1.50000 1.50000
17 B 1.4432068 3.9812135 0.5981310 1.50000 1.50000
18 B 5.4203018 3.6884584 4.4238291 1.50000 1.50000
19 B 6.3830427 7.0362738 -0.0827161 1.50000 1.50000
20 B 6.7972236 0.9258075 5.8321585 1.50000 1.50000
21 B 8.4061397 1.6141491 1.2538205 1.50000 1.50000
22 B 6.5651376 5.0168809 1.7394022 1.50000 1.50000
23 B 5.6819966 1.9668050 2.1926326 1.50000 1.50000
24 B 6.3651599 3.2655902 2.7294807 1.50000 1.50000
25 B -0.3300217 -1.3088082 0.8339259 1.50000 1.50000
26 B 0.8274979 3.4705711 2.7333313 1.50000 1.50000
27 B 2.9601606 3.5247227 -1.5252498 1.50000 1.50000
28 B 4.6117086 4.2872710 2.8702858 1.50000 1.50000
29 B 7.9118897 0.1295031 2.0848151 1.50000 1.50000
30 B 0.2254837 -4.6816994 2.4613908 1.50000 1.50000
31 B -0.8287363 6.9515258 1.9435684 1.50000 1.50000
32 B 1.5875014 -1.6825885 2.9574201 1.50000 1.50000
33 B 2.9186891 2.0413481 -0.6798196 1.50000 1.50000
34 B 3.5704564 -1.4377019 3.0431594 1.50000 1.50000
35 B 3.9155764 -0.4525063 1.4556058 1.50000 1.50000
36 B 1.0303640 -0.1362115 2.7688962 1.50000 1.50000
37 B -0.4224181 3.2727675 5.1581761 1.50000 1.50000
38 B 0.6583687 -2.3540293 5.3004636 1.50000 1.50000
39 B 1.4540576 0.7600311 4.2735441 1.50000 1.50000
40 B 6.2405750 1.0191593 0.1538643 1.50000 1.50000
41 B 5.1442027 0.9981257 6.3999104 1.50000 1.50000
42 B 1.8331635 4.8723317 -0.9630677 1.50000 1.50000
43 B -1.4692207 4.8554582 5.9627362 1.50000 1.50000
44 B 3.9415517 3.4679061 6.1783661 1.50000 1.50000
45 B 3.7475311 4.9931378 6.1835928 1.50000 1.50000
46 B 4.1650722 1.0000683 0.1391613 1.50000 1.50000
47 B 6.6501875 -0.9732179 6.1806047 1.50000 1.50000
48 B 2.0029135 2.6606199 1.6497126 1.50000 1.50000
49 B 6.5996274 6.9156290 4.0431469 1.50000 1.50000
50 B 3.4238033 5.6726295 -1.0725250 1.50000 1.50000
51 B 6.7725174 4.1210200 6.6073636 1.50000 1.50000
52 B 2.9701944 -1.2349994 -2.0587542 1.50000 1.50000
53 B -1.3495212 0.3169817 6.3110259 1.50000 1.50000
54 B 6.7603696 -0.4091677 -0.3385156 1.50000 1.50000
55 B 7.9395341 -0.6141338 -1.8684884 1.50000 1.50000
56 B 4.6080527 2.7198804 -1.9823130 1.50000 1.50000
57 B 5.3305896 6.2291412 5.1439166 1.50000 1.50000
58 B -0.4514767 5.1828337 0.6076977 1.50000 1.50000
59 B 8.2992850 -1.5809030 4.4091317 1.50000 1.50000
60 B 7.5960380 -2.8428759 0.7650555 1.50000 1.50000
61 B -0.9183885 -1.6655927 -3.5657607 1.50000 1.50000
62 B 6.1580067 -0.0352412 3.0132182 1.50000 1.50000
63 B 7.1477703 2.4957910 0.5359210 1.50000 1.50000
64 B 3.1106054 -0.3614047 4.5319334 1.50000 1.50000
65 B 1.3386776 -1.1207418 1.1694147 1.50000 1.50000
66 B -0.5676465 -1.1131825 2.9641196 1.50000 1.50000
67 B 5.9006825 -0.1860872 1.3520769 1.50000 1.50000
68 B 4.2652462 6.3764045 3.3771911 1.50000 1.50000
69 B -0.7669357 5.3302670 -0.7855952 1.50000 1.50000
70 B 2.1590092 4.9270463 4.8807802 1.50000 1.50000
71 B -0.3620619 4.9196476 4.0797430 1.50000 1.50000
72 B 7.2192710 6.6876886 6.4912873 1.50000 1.50000
73 B 1.1030345 5.0435666 6.3450783 1.50000 1.50000
74 B -0.1799631 6.1831397 -2.6760977 1.50000 1.50000
75 B 5.3150496 4.2419254 5.9013446 1.50000 1.50000
76 B 3.4165357 7.0144390 0.1456900 1.50000 1.50000
77 B 6.6526066 -0.7934385 4.4887465 1.50000 1.50000
78 B 2.2969623 -1.0518236 -0.3378298 1.50000 1.50000
79 B 6.3048644 -1.4211374 -1.8045264 1.50000 1.50000
80 B 0.5562409 6.2970971 -1.1100690 1.50000 1.50000
81 B 3.0540356 6.3890957 6.9146752 1.50000 1.50000
82 B 5.4238758 -2.0370411 6.6446369 1.50000 1.50000
83 B 5.2578858 1.2262796 -1.2662993 1.50000 1.50000
84 B 0.7793408 -1.7255644 -0.5353053 1.50000 1.50000
85 B 3.9337880 1.5529573 5.3755818 1.50000 1.50000
86 B 5.1354271 -1.7794068 1.1173890 1.50000 1.50000
87 B 7.8880935 -3.4015312 3.3374815 1.50000 1.50000
88 B -1.3396849 3.6961993 2.2866610 1.50000 1.50000
89 B 1.5361111 6.3456766 4.0869645 1.50000 1.50000
90 B 6.0092150 6.3529045 2.4139566 1.50000 1.50000
91 B 8.2679739 2.4522884 -0.7237842 1.50000 1.50000
92 B -0.6185959 1.0828455 -1.4292993 1.50000 1.50000
93 B 7.9664345 2.7913760 3.9766772 1.50000 1.50000
94 B 0.5182188 0.0790560 -0.3227246 1.50000 1.50000
95 B 8.0371402 3.7084658 -1.5452134 1.50000 1.50000
96 B 1.6665263 0.4181663 -3.7785911 1.50000 1.50000
Atoms.SpeciesAndCoordinates>
Atoms.UnitVectors.Unit Ang # Ang|AU
<Atoms.UnitVectors
9.51595 0.00000 0.00000
0.00000 9.51595 0.00000
0.00000 0.00000 9.51595
Atoms.UnitVectors>
#
# SCF or Electronic System
#
scf.XcType GGA-PBE # LDA|LSDA
scf.SpinPolarization off # On|Off
scf.ElectronicTemperature 700.0 # default=300 (K)
scf.energycutoff 150.0 # default=150 (Ry)
scf.maxIter 300 # default=40
scf.EigenvalueSolver Krylov # Recursion|Cluster|Band
scf.Kgrid 1 1 1 # means 4x4x4
scf.Mixing.Type rmm-diisk # Simple|Rmm-Diis|Gr-Pulay
scf.Init.Mixing.Weight 0.050 # default=0.30
scf.Min.Mixing.Weight 0.001 # default=0.001
scf.Max.Mixing.Weight 0.200 # default=0.40
scf.Mixing.History 7 # default=5
scf.Mixing.StartPulay 30 # default=6
scf.criterion 1.0e-6 # default=1.0e-6 (Hartree)
scf.lapack.dste dstevx # dstegr|dstedc|dstevx, default=dstegr
#
# 1D FFT
#
1DFFT.NumGridK 900 # default=900
1DFFT.NumGridR 900 # default=900
1DFFT.EnergyCutoff 3600.0 # default=3DFFT.EnergyCutoff*3.0 (Ry)
#
# Orbital Optimization
#
orbitalOpt.Method Off # Off|Unrestricted|Restricted
orbitalOpt.InitCoes Symmetrical # Symmetrical|Free
orbitalOpt.initPrefactor 0.1 # default=0.1
orbitalOpt.scf.maxIter 12 # default=12
orbitalOpt.MD.maxIter 2 # default=5
orbitalOpt.per.MDIter 2 # default=1000000
orbitalOpt.criterion 1.0e-4 # default=1.0e-4 (Hartree/borh)^2
#
# output of contracted orbitals
#
CntOrb.fileout off # on|off , default=off
Num.CntOrb.Atoms 2 # default=1
<Atoms.Cont.Orbitals
1
2
Atoms.Cont.Orbitals>
#
# SCF Order-N
#
orderN.HoppingRanges 6.0 # default=5.0 (Ang)
orderN.NumHoppings 2 # default=2
orderN.KrylovH.order 400
#
# MD or Geometry Optimization
#
MD.Type NVT_VS # Nomd|Constant_Energy_MD|Opt
# Constraint_Opt|DIIS2|Constraint_DIIS2
MD.maxIter 1000 # default=1
MD.TimeStep 1.0 # default=0.5 (fs)
MD.Opt.criterion 1.0e-4 # default=1.0e-4 (Hartree/bohr)
MD.Opt.DIIS_Mixing 0.1 # default=0.5
<MD.TempControl
1
1000 4 3000.0 0.8
MD.TempControl>
#
# MO output
#
MO.fileout off # on|off
num.HOMOs 1 # default=1
num.LUMOs 1 # default=1
MO.Nkpoint 1 # default=1
<MO.kpoint
0.0 0.0 0.0
MO.kpoint>
#
# DOS and LDOS
#
Dos.fileout off # on|off , default=off
Dos.Erange -10.0 10.0 # default = -20 20
Dos.Kgrid 9 9 9 # default = Kgrid1 Kgrid2 Kgrid3
#
# output Hamiltonian and overlap
#
HS.fileout off # on|off, default=off

I really appreciate your help.
With best regards,
Denis
メンテ
Re: problems with MD in OpenMX3.2 ( No.3 )
Date: 2007/04/13 10:18
Name: T.Ozaki

Hi,

Thank you for your input.
I tried your input file without any modification using OpenMX3.2,
and did not find any strange behaviour as shown below:

time= 0.000 (fs) Energy= -275.2221097 (Hatree) Temperature= 0.000
time= 1.000 (fs) Energy= -275.5708818 (Hatree) Temperature= 2731.9394027
time= 2.000 (fs) Energy= -275.2332310 (Hatree) Temperature= 2702.2627562
time= 3.000 (fs) Energy= -275.2441566 (Hatree) Temperature= 2697.5997913
time= 4.000 (fs) Energy= -275.2624816 (Hatree) Temperature= 2716.9871271

I am wondering why the total energy is so different from yours, and that
this is due to your specific problem or your machine specific problem.
Have you passed the 'runtest' on your machine ?

Also, I would have a couple of suggestions to your calculation as follows:

(1) Krylov or Band ?
Within the number of atoms in your system, the conventional 'Band' calculation
is faster than the 'Krylov'.

In fact the computational time for the first MD step was found to be

Band 137 (sec/MD step)
Krylov 278 (sec/MD step)

using 48 Opteron cpus. Thus, it is clear that the use of band is better.

(2) Basis function
For boron atom, B4.5-s2p2 will give reasonable results more than B6.5-s1p1.
B6.5-s1p1 must give a longer bond lengh more than the converged one.

Best regards,

TO
メンテ
Re: problems with MD in OpenMX3.2 ( No.4 )
Date: 2007/04/13 14:51
Name: Denis Music  <music@mch.rwth-aachen.de>

Hello,
Thanks a lot for these tips! Considering the runtest, with the 3.1 version, I've run CH4 and obtained exactly the same values as in the manual. I've also checked equilibrium volume and bulk moduli of diamond, Ge, alpha-boron, VO, alpha-alumina, etc. and I was pleased with the outcome. Furthermore, with OpenMX3.1 I have also obtained the values that you had obtained with OpenMX3.2 for the input file above, but OpenMX3.2 in may case gave the stange energy values. I'll do more tests with OpenMX3.2.
With best regards,
Denis
メンテ
Re: problems with MD in OpenMX3.2 ( No.5 )
Date: 2007/04/16 16:10
Name: Denis Music  <music@mch.rwth-aachen.de>

Hello,
I've reinstalled OpenMX3.2 with the patches and carried out the runtest. Here is the outcome:
1 input_example/Benzene.dat Elapsed time(s)= 19.93 diff Utot= 0.000000026173 diff Force= 0.000000000076
2 input_example/C60.dat Elapsed time(s)= 154.61 diff Utot= 0.000000153349 diff Force= 0.000000000841
3 input_example/CO.dat Elapsed time(s)= 102.09 diff Utot= 0.000000002951 diff Force= 0.001897890808
4 input_example/Cr2.dat Elapsed time(s)= 71.73 diff Utot= 0.000000014015 diff Force= 0.000000000053
5 input_example/Crys-MnO.dat Elapsed time(s)= 303.93 diff Utot= 0.000000015682 diff Force= 0.000000001268
6 input_example/GaAs.dat Elapsed time(s)= 345.61 diff Utot= 0.000000002324 diff Force= 0.000000000011
7 input_example/Glycine.dat Elapsed time(s)= 27.72 diff Utot= 0.000000001352 diff Force= 0.000000000027
8 input_example/Graphite4.dat Elapsed time(s)= 18.33 diff Utot= 0.000000002505 diff Force= 0.000000000060
9 input_example/H2O-EF.dat Elapsed time(s)= 32.23 diff Utot= 0.000000014494 diff Force= 0.000000000205
10 input_example/H2O.dat Elapsed time(s)= 20.24 diff Utot= 0.000000014604 diff Force= 0.000000000204
It doesn't look well, does it? I really do not understand why, since I didn't have any error messages while installing it.
Any suggestions?
With best regards,
Denis
メンテ
Re: problems with MD in OpenMX3.2 ( No.6 )
Date: 2007/05/14 18:49
Name: Denis Music  <music@mch.rwth-aachen.de>

Dear OpenMX users,
I'd like to inform you that the OpenMX 3.2 runs quite well on Opteron Solaris and Linux clusters. The latest patch does the trick for the Solaris Opteron version, while the Opteron Linux version works well once the “tee" command was removed. At the moment, I'm able to run with 156 atoms using the band method.
Greetings,
Denis
メンテ

Page: [1]