4
$\begingroup$

Ive been trying to do energy calculations under the influence of a perturbative field with the BLYP and B3LYP functionals. Using the following input files:

BLYP:

%mem=30000mb
%nproc=30
#P BLYP/aug-cc-pVQZ ExtraLinks=L608 SCF(CONVER=10,nosymm) NoSymmetry  field=read UNITS=BOHR

water

0 1
C   1.454760  -0.000000  0.000000
C   -1.454760  -0.000000  0.000000
H   2.207680  -0.000000  -1.930640
H   2.207560  -1.672010  0.965270
H   2.207720  1.672010  0.965270
H   -2.207720  -1.672010  -0.965280
H   -2.207560  1.672010  -0.965270
H   -2.207680  0.000000  1.930640

-0.006   -0.006   0.0
0.0  0.0  0.0
0.0  0.0  0.0

            

B3LYP:

%mem=30000mb
%nproc=30
#P B3LYP/aug-cc-pVQZ ExtraLinks=L608 SCF(CONVER=10,nosymm) NoSymmetry  field=read UNITS=BOHR

water

0 1
C   1.454760  -0.000000  0.000000
C   -1.454760  -0.000000  0.000000
H   2.207680  -0.000000  -1.930640
H   2.207560  -1.672010  0.965270
H   2.207720  1.672010  0.965270
H   -2.207720  -1.672010  -0.965280
H   -2.207560  1.672010  -0.965270
H   -2.207680  0.000000  1.930640

-0.006   -0.006   0.0
0.0  0.0  0.0
0.0  0.0  0.0

After Ive achieved normal termination for both these are the execution times:

BLYP:

Job cpu time:       3 days  7 hours 38 minutes  7.0 seconds.
Elapsed time:       0 days  2 hours 39 minutes 41.4 seconds.

B3LYP:

 Job cpu time:       0 days  0 hours 39 minutes 54.9 seconds.
 Elapsed time:       0 days  0 hours  1 minutes 20.6 seconds.

By comparing the Log files: The ICntrl variable was set to 0 for B3LYP and 500 for BLYP, which seems to be whats making the B3LYP calculations faster. Given below are the excerpts from the log files.

BLYP:

 FoFCou: FMM=F IPFlag=           0 FMFlag=      100000 FMFlg1=           0
         NFxFlg=           0 DoJE=T BraDBF=F KetDBF=T FulRan=T
         wScrn=  0.000000 ICntrl=       500 IOpCl=  0 I1Cent=   200000004 NGrid=           0
         NMat0=    1 NMatS0=      1 NMatT0=    0 NMatD0=    1 NMtDS0=    0 NMtDT0=    0
 Symmetry not used in FoFCou.
 Harris En= -79.9290251867264    

B3LYP:

FoFCou: FMM=F IPFlag=           0 FMFlag=      100000 FMFlg1=           0
         NFxFlg=           0 DoJE=F BraDBF=F KetDBF=F FulRan=T
         wScrn=  0.000000 ICntrl=         0 IOpCl=  0 I1Cent=           0 NGrid=           0
         NMat0=    1 NMatS0=      1 NMatT0=    0 NMatD0=    1 NMtDS0=    0 NMtDT0=    0
 Symmetry not used in FoFCou.

Is there anyway to manually set the ICntrl variable to zero?

Note: The reason for suspecting the ICntrl variable in the log file is that, the B3LYP calculations run slower than BLYP (as expected) when I use the aug-cc-pVTZ basis set. And the ICntrl is set to 500 in both log files.

$\endgroup$
3
  • $\begingroup$ Is ICntrl specifically the same for this section FoFCou when running with aug-cc-pvtz? My understanding is ICntrl is a name that is reused with various meanings for different routines. As an example see IOp 1/75 here, where ICntrl is mentioned in an entirely different context. $\endgroup$
    – Tyberius
    Commented May 6 at 1:36
  • $\begingroup$ @Tyberius Ive tried to use IOp(1/75=0) to set it to zero. But it still shows a non zero value as mentioned in the question. So I'm not sure how to proceed further $\endgroup$
    – Atom
    Commented May 9 at 21:41
  • $\begingroup$ What I'm saying is that ICntrl is likely an entirely internal variable, that changes meaning and value throughout Gaussian. So I suspect it's a variable that in general can't be set by the user. $\endgroup$
    – Tyberius
    Commented May 9 at 23:24

0

You must log in to answer this question.