From owner-chemistry@ccl.net Fri Aug 18 01:53:00 2006 From: "andras.borosy[]givaudan.com" To: CCL Subject: CCL: 1 CPU hour at IBM SP2 from 2000? Message-Id: <-32405-060817130119-8922-CHX/7O48sygNDgN7PGo5Bg+/-server.ccl.net> X-Original-From: andras.borosy],[givaudan.com Content-Type: multipart/alternative; boundary="=_alternative 005D7E0CC12571CD_=" Date: Thu, 17 Aug 2006 19:01:10 +0200 MIME-Version: 1.0 Sent to CCL by: andras.borosy[*]givaudan.com This is a multipart message in MIME format. --=_alternative 005D7E0CC12571CD_= Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable Dear Colleagues, How many, say, Athlon 2.4GHz CPU time is equivalent to 1 CPU hour at IBM = SP2 from 2000? Many thank, Dr. Andr=E1s P=E9ter Borosy Senior Scientist Fragrance Research Givaudan Schweiz AG - Ueberlandstrasse 138 - CH-8600 - D=FCbendorf -= =20 Switzerland T:+41-1-824 2164 - F:+41-1-8242926 - http://www.givaudan.com --=_alternative 005D7E0CC12571CD_= Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable
Dear Colleagues,

How many, say, Athlon 2.4GHz CPU time is equivalent  to 1 CPU hour at  IBM SP2 from 2000?

Many thank,

Dr. Andr=E1s P=E9ter Borosy
Senior Scientist
Fragrance Research
Givaudan Schweiz AG  -  Ueberlandstrasse 138  -  CH-8600  -  D=FCbendorf  -  Switzerland
T:+41-1-824 2164  -  F:+41-1-8242926    -  http://= www.givaudan.com --=_alternative 005D7E0CC12571CD_=-- From owner-chemistry@ccl.net Fri Aug 18 09:45:00 2006 From: "Mikael Johansson mikael.johansson!=!helsinki.fi" To: CCL Subject: CCL: Combining Turbomole fragment orbitals Message-Id: <-32406-060818093112-21802-VNC8NBRcYCNG1/5h8oTS9w- -server.ccl.net> X-Original-From: Mikael Johansson Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Date: Fri, 18 Aug 2006 15:20:02 +0300 (EEST) MIME-Version: 1.0 Sent to CCL by: Mikael Johansson [mikael.johansson * helsinki.fi] Hello All! Inspired by a mail on CCL yesterday, I wrote a small script that combines the orbitals of two systems (fragments) into one for Turbomole. It's called moscomb and available for the interested here: http://www.iki.fi/~mpjohans/scripts/ It's not much more than a copy/paste script in Python (quite an ugly one at that). It would of course have been much better to actually read in the ao coefficients as real data objects, but I thought a simple copy/paste would be faster to do. In hindsight, I think I was wrong... Oh well, perhaps when the need to do some more manipulation on the orbitals arises, I'll get around to do it properly. Anyway, I guess it's better than nothing. I've tested it on a few systems, and it seems to work OK. For a combined system of one low-spin and one high-spin haem, the 2nd SCF-iteration energy is > 20 Hartree lower in energy when starting from converged fragment mo's compared to EHT-orbs. Also, the crazy oscillations are reduced significantly. I would be quite surprised if it would be bug free, though :-) It works (should work) only for C1 symmetry. If your combined system has higher symmetry, too bad. If your fragments have higher symmetry, you can of course calculate those in whatever symmetry and then use for example define to transform them into C1. When combining the systems, the atoms need to be in the same order as in the fragments, with fragment 2 directly following fragment 1. Have a nice day, Mikael J. http://www.iki.fi/~mpjohans From owner-chemistry@ccl.net Fri Aug 18 10:31:01 2006 From: "Lubos Vrbka lists=-=vrbka.net" To: CCL Subject: CCL:G: gausian98 and nbo 5.0 on solaris Message-Id: <-32407-060818094126-22536-dK6ZOgXrKFYtM/HQwwWu6Q|*|server.ccl.net> X-Original-From: Lubos Vrbka Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-2; format=flowed Date: Fri, 18 Aug 2006 14:45:17 +0200 MIME-Version: 1.0 Sent to CCL by: Lubos Vrbka [lists.-$-.vrbka.net] hi guys, i'm trying to get g98 + nbo running together on a solaris machine. i'm however not able to compile it. since i don't have almost any experience with solaris, i want to ask for a help here, in case somebody was able to proceed. the compilation goes on ok. however, with l607; ... a - unpack.o a - valdm.o a - valtbl.o a - veclen.o a - wgtopt.o a - wgtout.o a - wgtpr.o a - wnorm.o a - worth.o a - wortun.o a - wrarc.o a - wrbas.o a - wrlbl.o a - wrnlmo.o a - wrppna.o a - wrsp.o a - wrtnab.o a - wrtnao.o a - wrtnbo.o a - xcited.o rm -f -r temp-l607 bsd/fixlib l607.a ar x main.a ml607.o f95 -Bstatic -L/workshop62/opt/SUNWspro/WS6U2/lib/v9 -lfsu -lfui -lifai -lfai -lfai2 -fast -xtypemap=real:64,double:64,integer:64 -xtarget=native -xarch=v9 -xcache=generic -o l607.exe ml607.o l607.a util.a -xlic_lib=sunperf -Bdynamic -lc -lm -fast -xtypemap=real:64,double:64,integer:64 -xtarget=native -xarch=v9 -xcache=generic ld: fatal: symbol `order_' is multiply-defined: (file l607.a(order.o) type=FUNC; file util.a(redofs.o) type=OBJT); ld: fatal: File processing errors. No output written to l607.exe *** Error code 1 make: Fatal error: Command failed for target `l607.exe' endif any hints how to proceed further? i'm trying to get it working because the g03+nbo 5g (as written in the manual) doesn't support nmr calculation keyword NCS (although it seems to work, however the numbers obtained do not agree with the numbers printed in the manual...) thank you in advance for any hints. with best regards, -- Lubos _-$-_" http://www.lubos.vrbka.net From owner-chemistry@ccl.net Fri Aug 18 12:46:00 2006 From: "S.I.Gorelsky gorelsky-x-stanford.edu" To: CCL Subject: CCL: combining fragment wave functions Message-Id: <-32408-060818124416-24674-i9RZM/+66KUIOqYenAwodQ:server.ccl.net> X-Original-From: "S.I.Gorelsky" Content-Type: TEXT/PLAIN; charset=US-ASCII Date: Fri, 18 Aug 2006 09:24:43 -0700 (PDT) MIME-Version: 1.0 Sent to CCL by: "S.I.Gorelsky" [gorelsky * stanford.edu] Combining fragment wave functions to generate a guess wave function for a molecular system that is a sum of many "chemically-well-defined" fragments is, in most situations, a MUCH BETTER GUESS than the default "atomic" initial guess methods. Unfortunately, the method has been underutilized. I think it is time to remedy that. The most useful application of this method is not for two-fragment systems but for systems of weakly, non-covalently coupled molecules with 100s, 1000s of atoms where a user will "pull out" converged wave functions of fragments from the database to generate a guess wave function of the whole system. This is how AOMix-CDA (http://www.sg-chem.net/aomix-s/) will handle this task. Serge Gorelsky http://www.sg-chem.net ----------------------------------------------------------------- Dr. S.I. Gorelsky, Department of Chemistry, Stanford University Box 155, 333 Campus Drive, Stanford, CA 94305-5080 USA Fax: (650) 960-3912 ----------------------------------------------------------------- From owner-chemistry@ccl.net Fri Aug 18 13:28:00 2006 From: "Luis M Simon luissimonrubio-*-hotmail.com" To: CCL Subject: CCL: AIMPAC compilation Message-Id: <-32409-060818131608-11747-cyRsZ4MLWnOqsFuH6qNW9w,+,server.ccl.net> X-Original-From: "Luis M Simon" Date: Fri, 18 Aug 2006 13:16:08 -0400 Sent to CCL by: "Luis M Simon" [luissimonrubio:hotmail.com] Dear CCLers: I suppouse that someone else have already raised that question tot he list, but after a brief search in google I have not found a solution to my problem. It is very simple: How do I install in linux the AIMPAC software (I am particularly interested in BUBBLE module). In the web page where I downoloaded AIMPAC I found instructions only to intall graphical modules grdvec and contor. I installed the pgplot libraries and compile both files using the compilepg.csh script. Nevertheless, I have not been able to install any of the other probrams included in the AIMPAC distribution. The .SRC files appears to be fortran files, so I tried g77 -o bubble BIBBLE.SRC. It raises an error with a message of unrecognized language. So I tried to rename BIBBLE.SRC to BIBBLE.f and then I obtained lots errors saying that common blocks has different lenght (220 bytes and 120 bytes) in different parts of the code... (I have checked that there are two different deffinitions of the same common block in different part of the code, and it is now allowed by the compiler) Is it a compiler problem? Do I need some makefile that is not in the distribution? Do I have to use any scpecific compiler option? Or do I need a fortran90 compiler? I am sorry because I am not experienced at all with compiling and I hope that this is a very simple problem. Thank you for your help. Luis Simon