[molpro-user] DFSAPT exch-disp calculation fails

[Enrique Cabaleiro]

Dear Molpro

I am having some problems with SAPT(DFT) calculations in large systems
using the new DFSAPT module in Molpro15.

I have been able to run a job with 2328 basis functions and 168 electrons
without problem, just asking for MEMORY,600,M.

However, when I increase the size of the system to 2372 basis functions and
198 electrons the job always stops  with a segmentation fault message, just
before ending the calculation of exch-dispersion contribution. I have
increased the memory to MEMORY,1200,M and even to MEMORY,7000,M with no
result.
In all cases the program stops before ending the Exch-disp calculation:

 E(2)exch-disp
 =============
 (ia|jb).Sib/Sja...                           CPU time:   107.51
 (P|ia).SSi transformed integrals...          CPU time:   490.12
 (P|jb).SSj transformed integrals...          CPU time:   323.79
 (P|ia).Sij transformed integrals...          CPU time:   436.02
 (P|jb).Sji transformed integrals...          CPU time:   452.26
 done:     2.74%    CPU time:  1692.69
 done:     5.48%    CPU time:  1619.34
 done:     8.22%    CPU time:  1626.85
 done:    10.96%    CPU time:  1574.67
 done:    13.70%    CPU time:  1584.20
 done:    16.44%    CPU time:  1433.22
 done:    19.18%    CPU time:  1337.00
 done:    21.92%    CPU time:  1284.13
 done:    24.66%    CPU time:  1530.77
 done:    27.40%    CPU time:  1619.63
 done:    30.14%    CPU time:  1577.61
 done:    32.88%    CPU time:  1620.97
 done:    35.62%    CPU time:  1580.66
 done:    38.36%    CPU time:  1615.58
 done:    41.10%    CPU time:  1563.07
 done:    43.84%    CPU time:  1582.54
 done:    46.58%    CPU time:  1400.51
 done:    49.32%    CPU time:  1306.42
 done:    52.05%    CPU time:  1561.26
 done:    54.79%    CPU time:  1538.78
 done:    57.53%    CPU time:  1567.66
 done:    60.27%    CPU time:  1296.91
 done:    63.01%    CPU time:  1271.76
 done:    65.75%    CPU time:  1273.03
 done:    68.49%    CPU time:  1272.98
 done:    71.23%    CPU time:  1276.71
 done:    73.97%    CPU time:  1277.00
 done:    76.71%    CPU time:  1277.53
 done:    79.45%    CPU time:  1278.63
 done:    82.19%    CPU time:  1270.43
 done:    84.93%    CPU time:  1214.99
 done:    87.67%    CPU time:  1215.03
 done:    90.41%    CPU time:  1217.58
 done:    93.15%    CPU time:  1214.26
 done:    95.89%    CPU time:  1216.07
 done:    98.63%    CPU time:  1215.97


I have seen in JCP 141, 094107 (2014), that much larger calculations can be
carried out, so maybe I am misinterpreting something.

This is my input file:


MEMORY,1200,M
gdirect; gthresh,THROVL=1.e-9
angstrom
geomtyp=xyz
symmetry,nosym orient,noorient
geometry={
include, ghexa-hexa.xyz
}


basis={
set,orbital; default,vtz
set,jkfit;   default,vtz/jkfit
set,mp2fit;  default,vtz/mp2fit
set,dflhf;   default,vtz/jkfit
}


!================================================================
ca=2101.2; cb=2102.2
oldnorm=1
ipghexa=0.30481511
iphexa=0.21937674
hoghexa=-0.263564
hohexa=-0.177575
deltaA=ipghexa+hoghexa
deltaB=iphexa+hohexa
!================================================================

dummy,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,
87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,
105,106,107,108,109,110
{df-ks,lhf,pbex,pbec,basis=jkfit; dftfac,0.25,0.75,1.0; asymp,deltaA;
save,$ca; start,atdens}
monomerA,orb=$ca,core=26

dummy,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,
22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,
42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,
62,63,64,65,66,67,68
{df-ks,lhf,pbex,pbec,basis=jkfit; dftfac,0.25,0.75,1.0; asymp,deltaB;
save,$cb; start,atdens}
monomerB,orb=$cb,core=26

{dfsapt,all=1,auxbas='mp2fit',fcoul=1.0,fxc=1.0,fxcfit=1,fxcsing=1,t678=3}


Thanks,
Quique

-- 
Enrique M. Cabaleiro Lago
Departamento de Química Física
Facultad de Ciencias-Lugo
Universidad de Santiago de Compostela
Tel: 982295800 Ext. 24129
Fax: 982285872
e-mail: caba.lago at usc.es
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20161109/36bfce54/attachment.html>