[CPMD-list] [Fwd: Fw: VDB on BlueGene]

[Axel Kohlmeyer]

On Mon, 23 Jul 2007, Philip Shemella wrote:


PS> Dear CPMD'ers,  Forgive the possible double post.
PS> 
PS> Brock is running into some problems with the B_VDB_PBE_NLCC.psp and 
PS> corrected Al_VDB_PBE_NLCC.psp.    I think the error is actually not 
PS> related to blue gene because it was reproduced on an AMD cluster.  The 
PS> B_VDB_PBE_NLCC.psp works for BH3, a B atom, and other test cases.

phil, brock
please note, that the error message is pretty clear:
there is no TASKGROUP support for VDB pseudopotentials.
that will happen on any machine. the potentials should
work without TASKGROUPs, though.

you should also be able to generate a matching norm-conserving
boron pseudopotential from the existing boron BLYP potential
(just use the same cutoffs). boron is not a very critical
element to generate a pseudopotential for.

cheers,
   axel. 

PS> 
PS> Thanks, Phil Shemella
PS> RPI
PS> 
PS> ==============Original message text===============
PS> From: Brock Russell <russeb4 at rpi.edu>
PS> Date: Mon, 23 Jul 2007 9:03:43 EDT
PS> Subject: VDB on BlueGene
PS> 
PS> Hello.
PS> 
PS> I am attempting to run a geometry optimization of a silicon and boron 
PS> system on Blue Gene.  The only PBE pseudopotential for boron that I have 
PS> is VDB.  I have not been able to successfully run the optimization, 
PS> however.  I get the error message "PROGRAM STOPS IN SUBROUTINE RHOV| NO 
PS> GROUPS ALLOWED [PROC= 120]".  Has anyone else encountered this problem 
PS> and found a solution?  Below is my input file.  Thank you.
PS> 
PS> &CPMD
PS>    OPTIMIZE GEOMETRY
PS>    CONVERGENCE ORBITALS
PS>     1.0d-6
PS>    CONVERGENCE GEOMETRY
PS>     1.0d-3
PS>    MIRROR
PS>    REAL SPACE WFN KEEP
PS>    MEMORY BIG
PS>    TASKGROUPS
PS>    4
PS>    LSD
PS> &END
PS> 
PS> &SYSTEM
PS>     ANGSTROM
PS>     SYMMETRY
PS>     8
PS>     CELL
PS>     15.44604 1.29483 1.29483 0 0 0
PS>     CUTOFF
PS>     60.0
PS>     MULTIPLICITY
PS>     2
PS> &END
PS> 
PS> &DFT
PS>    NEWCODE
PS>    FUNCTIONAL PBE
PS> &END
PS> 
PS> &ATOMS
PS> *H_MT_PBE.psp KLEINMAN-BYLANDER
PS>    LMAX=S
PS>     48
PS>      -0.601739631344      6.087628504715     -5.542406785273
PS>      -2.646887735823      4.871738535216     -5.556326330436
PS> ...
PS>       8.811454620704      1.560681735720    -10.632451505536
PS>      12.672684841313      1.562379491434    -10.627786212344
PS> 
PS> *Si_MT_PBE.psp KLEINMAN-BYLANDER
PS>    LMAX=D
PS>     63
PS>      -0.851095059925     -0.670949347743     -1.704606351653
PS>       1.116030039990     -3.399230179248     -3.641686059825
PS>       1.104390383716      0.682368765990     -5.562235088354
PS> ...
PS>      10.433930271062      4.703038028393     -5.564567861554
PS> 
PS> *B_VDB_PBE_NLCC.psp FORMATTED NEWF NLCC
PS>    LMAX=P
PS>    1
PS>     -0.815283967236     -2.044258752853     -3.619975649824
PS> 
PS> &END
PS> 
PS> Brock
PS> RPI
PS> 
PS> 
PS> ===========End of original message text===========
PS> 
PS> 
PS> 
PS> 
PS> 
PS> 
PS> 
PS> 
PS> _______________________________________________
PS> CPMD-list mailing list
PS> CPMD-list at cpmd.org
PS> http://cpmd.org/mailman/listinfo/cpmd-list
PS> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.