<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>Thank you for your reply,<br>I wasn't sure what things I did before, exactly, so I recompiled it as serial version without doing anything extra, last week - the problem is not the parallelism (the openmp part was, to be able to use the parallel mkl-libraries - I have some general questions there, but I'll ask them later, independent of this topic).<br><br>This time I configured with:<br>./configure -prefix $insdirpath -blas -lapack -icc -ifort -blaspath /path/intel/mkl111/lib/intel64 -lapackpath /path/intel/mkl111/lib/intel64<br>and used the default for every question asked, thereafter. (This, btw. results in not having a lapackpath set, there isn't even a question for this, and the entry in CONFIG is empty.)<br><br>I used ifort 14.0 now, as it's listed as supported for linux x86_64, and again the mkl 11.1 (which is not, but from prior tests I know, is quite a bit faster). The compilation runs through normally and every test passes, except, again, the three fciqmc ones. And again I get the gradient error.<br>I'll attach the CONFIG file, the errout of be2_odds_fciqmc and the out-file of my he-chain.<br><br>I just compiled it without using mkl (so just ./configure -prefix $insdirpath -icc -ifort and hitting enter for all questions), which also gave a binary that ran the first testjob fine (I didn't do the whole suite, now), but failed to run be2_odds_fciqmc.<br>Sven<br><br><br><div>> Date: Wed, 16 Oct 2013 10:02:36 +0100<br>> From: MayAJ1@cardiff.ac.uk<br>> To: goretoffel@hotmail.de; molpro-user@molpro.net<br>> Subject: Re: [molpro-user] compiling problems/bug: multi-cpmcscf gradients with >50 atoms<br>> <br>> Sven,<br>> <br>> Is there any reason you have changed FFLAGS and CFLAGS? We only use the <br>> Fortran bindings of BLAS/LAPACK so I don't think the the -I options, or <br>> -DMKL_ILP64 will make any difference, and -i8 should be default on <br>> 64-bit anyway. The only flags I think are making a difference are the <br>> -openmp flags, although it is more convenient to just specify -openmp <br>> directly to configure and let it do the rest. I would suggest to try <br>> building without openmp since there is hardly any openmp code in Molpro, <br>> certainly the binaries are built without openmp.<br>> <br>> Best wishes,<br>> <br>> Andy<br>> <br>> On 15/10/13 14:10, Sve N wrote:<br>> > Dear molpro users and developers,<br>> > when compiling molpro myself, it finishes all testjobs of 'make test',<br>> > except the three fciqmc ones, properly (the fciqmc jobs finish running,<br>> > too, but end up giving too large errors in the energy, something like<br>> > 0.0006). Since I do not plan on using it, it didn't seem important to<br>> > me. My most recent configure-options (all of this mail holds true for<br>> > patch 0, 1 and 6 with various options I tried) are:<br>> ><br>> > ./configure -prefix $insdirpath -blas -lapack -icc -ifort -mpp<br>> > -auto-ga-mvapich2 -var 'BLASLIB=-Wl,--start-group<br>> > /path/intel/mkl111/lib/intel64/libmkl_intel_ilp64.a<br>> > /path/intel/mkl111/lib/intel64/libmkl_core.a<br>> > /path/intel/mkl111/lib/intel64/libmkl_intel_thread.a -Wl,--end-group<br>> > -lpthread -lm' -var 'LAPACKLIB=-Wl,--start-group<br>> > /path/intel/mkl111/lib/intel64/libmkl_intel_ilp64.a<br>> > /path/intel/mkl111/lib/intel64/libmkl_core.a<br>> > /path/intel/mkl111/lib/intel64/libmkl_intel_thread.a -Wl,--end-group<br>> > -lpthread -lm' -var 'CFLAGS=-DMKL_ILP64 -openmp<br>> > -I/path/intel/mkl111/include' -var 'FFLAGS=-i8 -openmp<br>> > -I/path/intel/mkl111/include'<br>> ><br>> > The compilers are ifort and icc 13.1, mkl's version is 11.1. The two<br>> > cpus I let it run on, are an Intel Xeon E5-2630L and an Intel Xeon<br>> > L5630, although it is compiled on an AMD Opteron 6128 HE (is that a<br>> > problem?).<br>> ><br>> > When doing a gradient calculation in a multi-block with more than 50<br>> > atoms, I get wrong gradients for the atoms after the 50th. This doesn't<br>> > happen with the prebuilt binaries, so it has to be a compiling problem.<br>> > I have no clue, how to fix this, so if you have any advice, I'd be thankful.<br>> ><br>> > As an example I calculated a He-chain, with the He-atoms seperated by 3<br>> > Angstrom, and little deviations in the y-direction.<br>> ><br>> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br>> > ***<br>> > memory,50,m<br>> ><br>> > basis<br>> > default,6-31g<br>> > end<br>> ><br>> > angstrom<br>> > nosym<br>> > orient,noorient<br>> > geomtyp=xyz<br>> > geometry={<br>> > 52<br>> ><br>> > he 0. 0. 0.<br>> > he 3. 1. 0.<br>> > he 6. 0. 0.<br>> > [...]<br>> > he 147. 1. 0.<br>> > he 150. 0. 0.<br>> > he 153. 0. 0.<br>> > }<br>> ><br>> > hf<br>> ><br>> > force<br>> ><br>> > {multi<br>> > closed,51<br>> > occ,53<br>> > state,2<br>> > cpmcscf,grad,1.1<br>> > }<br>> ><br>> > force<br>> ><br>> > ---<br>> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br>> ><br>> > This results in (regardless of the number of processors used):<br>> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br>> > 1PROGRAM * RHF-SCF (CLOSED SHELL) Authors: W. Meyer, H.-J. Werner<br>> ><br>> ><br>> > NUMBER OF ELECTRONS: 52+ 52- SPACE SYMMETRY=1 SPIN<br>> > SYMMETRY: Singlet<br>> > [...]<br>> > 1PROGRAM * ALASKA (Gradient of the energy) Author: R. Lindh<br>> > [...]<br>> > SCF GRADIENT FOR STATE 1.1<br>> ><br>> > Atom dE/dx dE/dy dE/dz<br>> ><br>> > 1 0.000002476 0.000000825 0.000000000<br>> > 2 -0.000000001 -0.000001650 0.000000000<br>> > [...]<br>> > 49 -0.000016545 0.000000825 0.000000000<br>> > 50 -0.000000000 -0.000001650 0.000000000<br>> ><br>> > 51 0.000016546 0.000000825 0.000000000<br>> > 52 -0.000019022 0.000000000 0.000000000<br>> > [...]<br>> > 1PROGRAM * MULTI (Direct Multiconfiguration SCF) Authors: P.J.<br>> > Knowles, H.-J. Werner (1984) S.T. Elbert (1988)<br>> > [...]<br>> > 1PROGRAM * FORCE (Gradient of the energy)<br>> ><br>> > SA-MCSCF information from 5100.1<br>> > [...]<br>> > SA-MC GRADIENT FOR STATE 1.1<br>> ><br>> > Atom dE/dx dE/dy dE/dz<br>> ><br>> > 1 0.000002476 0.000000825 0.000000000<br>> > 2 -0.000000001 -0.000001650 0.000000000<br>> > [...]<br>> > 49 -0.000016545 0.000000825 0.000000000<br>> > 50 0.000000000 -0.000001650 0.000000000<br>> ><br>> > 51 0.000016546 0.000000825 289.128098107<br>> > 52 -0.000019022 0.000000000 289.128098107<br>> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br>> > (The blank line between atom 50 and 51 appears in the out-file, like<br>> > it's shown here.)<br>> ><br>> > Thanks, Sven<br>> ><br>> ><br>> > _______________________________________________<br>> > Molpro-user mailing list<br>> > Molpro-user@molpro.net<br>> > http://www.molpro.net/mailman/listinfo/molpro-user<br>> ><br></div> </div></body>
</html>