[molpro-user] problem of parallel MCSCF calculation
Peter Knowles
KnowlesPJ at Cardiff.ac.uk
Tue Jul 14 19:49:08 BST 2009
This looks to me like the same problem as in http://www.molpro.net/pipermail/molpro-user/2009-July/003171.html
. If you have fully-patched source code for Molpro 2008.1, or the
latest Mac binary, you should not see the problem; unfortunately the
supplied linux binaries do not yet include the fix, but will soon.
Peter
On 14 Jul 2009, at 18:00, Jingjing Zheng wrote:
> Dear All,
>
> I met a problem when I run a MCSCF calculation using parallel code of
> Molpro both version 2006.1(patch 148) and 2008 (patch 5). It can run
> without problem using the serial code.
>
> The job crashed after entering MULTI code, and here is the error
> message. I found the similar problem in the mail list, but there is no
> answer there. Any suggestions are appreciated!
>
> ITER. MIC NCI NEG ENERGY(VAR) ENERGY(PROJ) ENERGY
> CHANGE GRAD(0) GRAD(ORB) GRAD(CI) STEP TIME
>
> USED MEMORY IN cislow: 4813 18383 18383
> 18383 18383 18383 18383 18383 18383 18383
> FREE MEMORY IN cislow:
> 1099995087109998151710999815171099981517109998151710999815171099981517109998151710999815171099981517
> ? Error
> ? Inconsistent memory
> ? The problem occurs in check_address
>
> GLOBAL ERROR fehler on processor 0
>
>
> My input file is:
>
> ***,butoxy
> memory,400,M
> gprint,basis,orbitals
> geomtyp=xyz
> geometry={
> nosym;noorient;
> 14
>
> C , -1.384361 , 0.403657 , -0.009933
> C , -0.084521 , -0.385233 , -0.005979
> C , 1.138686 , 0.523850 , 0.006073
> C , 2.445775 , -0.259679 , 0.000215
> O , -2.527285 , -0.343813 , 0.004909
> H , -1.447091 , 1.060205 , 0.877246
> H , -1.432124 , 1.105553 , -0.858653
> H , -0.075378 , -1.042750 , 0.866209
> H , -0.062431 , -1.029494 , -0.888118
> H , 1.104962 , 1.188409 , -0.861553
> H , 1.100983 , 1.169702 , 0.887687
> H , 3.310094 , 0.402925 , 0.009744
> H , 2.510880 , -0.909978 , 0.873214
> H , 2.514911 , -0.889633 , -0.887294
> }
>
> basis=vdz
> {hf;wf,41,1,1;}
> {multi;occ,22;frozen,0;closed,19;
> wf,41,1,1;maxiter,20;
> rotate, 17.1, 20.1;
> rotate, 22.1, 49.1}
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
--
Prof. Peter J. Knowles
School of Chemistry, Cardiff University, Main Building, Park Place,
Cardiff CF10 3AT, UK
Email KnowlesPJ at Cardiff.ac.uk
WWW http://www.cardiff.ac.uk/chemy/contactsandpeople/academicstaff/knowles.html
More information about the Molpro-user
mailing list