alaska parallelization...


[ Molcas user's WWWBoard ]

Posted by Joubert Doriol Loïc on March 11, 2011 at 14:10:23:

Hello all.
I want to optimize the geometry at the MS-CASPT2 level of theory.
I use numerical calculation of gradients because analytic's are not implemented.
As this kind of calculation takes several days or weeks in serial, I am trying to use the parallel version of MOLCAS especially for the alaska part which seems to be well parallelized.

The problem is that I cannot find any keyword saying I want to run a parallel calculation. Indeed, when I use the following input, I observe a serial gradient calculation in the output file (I indicate the number of processor in the CPUS variable then I export it).

&GATEWAY &END
coord
$Project.xyz
group = C1
basis = ANO-S-VDZP
End of input
>>>>>>>>>>>>> Do while <<<<<<<<<<<<
&SEWARD &END
CHOLesky
End of input
&RASSCF &END
NACTel
12 0 0
CIROot
2 4
1 2
1 1
INACtive
29
RAS2
11
End of input
&CASPT2 &END
MULTistate
2 1 2
End of ifile
&Alaska &END
&SLAPAF &END
>>>>>>>>>>>>> ENDDO <<<<<<<<<<<<<<


Does someone know what is the problem?
Sorry if I just missed the keywords I should use when I read the manual.

Thanks


Follow Ups:



Post a Followup

Name:
E-Mail:

Subject:

if B is 1s22s22p1, what is Li?

Passfield:

Comments:


[ Follow Ups ] [ Post Followup ] [ Molcas user's WWWBoard ]