From airy at moegi.waseda.jp Fri Apr 1 10:36:39 2005 From: airy at moegi.waseda.jp (Junichi Kato) Date: Fri, 1 Apr 2005 17:36:39 +0900 Subject: [CPMD-list] Visualization of CPMD Message-ID: <006401c53695$ecb30770$8708a8c0@T42> Hi. I begin vizualization of CPMD with Axel home page. Though this page is easy to understand, I have struggle with secion 3.3 titled Extending the Script and Plugin Search Path. I have installed VMD1.8.3 Windows version by the defalut option. Where should I locate vmd.rc for being additional Tcl scripts (=subroutines) and plugins avaliable for Windows? I have get the error that Couldnt found additional Tcl scripts and plugins. What is worng? I have added the scentence of the following in vmd.rc and located vmd.rc in My Document directory. //////////////////////////////////////////////////////////////////////# search for new/updated molfile plugins in $HOME/vmd/plugins/$VMDARCH/molfile # type 'vmdinfo arch' to find your name for $VMDARCH vmd_plugin_scandirectory [file join $env(HOME) vmd/plugins [vmdinfo arch] molfile] *.so # add local (autoloaded) scripts to the search path set auto_path [concat $env(HOME)/vmd/scripts/tcl $auto_path] # command extensions (e.g. vmd_draw_vector) have to be sourced foreach ext [glob -nocomplain $env(HOME)/vmd/scripts/extensions/*.tcl ] { source $ext } unset ext ////////////////////////////////////////////////////////////////////// Then made directories named vmd/scripts, vmd/scripts/tcl, vnd/scripts/extensions, and plugins, respectively in My Document directory, and moved addtional files to those directories. Junichi kato -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050401/afb744ff/attachment.html From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Fri Apr 1 11:58:28 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Fri, 1 Apr 2005 11:58:28 +0200 (CEST) Subject: [CPMD-list] Visualization of CPMD In-Reply-To: <006401c53695$ecb30770$8708a8c0@T42> Message-ID: On Fri, 1 Apr 2005, Junichi Kato wrote: JK> Hi. hi! JK> I begin vizualization of CPMD with Axel home page. JK> Though this page is easy to understand, I have struggle with secion 3.3 JK> titled Extending the Script and Plugin Search Path. JK> I have installed VMD1.8.3 Windows version by the defalut option. JK> Where should I locate vmd.rc for being additional Tcl scripts JK> (=subroutines) and plugins avaliable for Windows? I have get the JK> error that Couldnt found additional Tcl scripts and plugins. What is JK> worng? well, this is more a VMD than a CPMD related question. since i don't use VMD on windows for anything serious (in fact, i very rarely use windows at all), there is little support for and no testing with my VMD scripts done on windows machines. i'd be happy to include additional comments for using VMD on Windows with my scripts, though. for your immediate question, please have a look at the VMD user's guide: http://www.ks.uiuc.edu/Research/vmd/current/ug/node193.html if you still have problems, please quote the exact error message and the corresponding section of your vmd.rc file. JK> I have added the scentence of the following in vmd.rc and located vmd.rc in JK> My Document directory. JK> //////////////////////////////////////////////////////////////////////# JK> search for new/updated molfile plugins in $HOME/vmd/plugins/$VMDARCH/molfile JK> # type 'vmdinfo arch' to find your name for $VMDARCH JK> vmd_plugin_scandirectory [file join $env(HOME) vmd/plugins [vmdinfo arch] JK> molfile] *.so JK> JK> # add local (autoloaded) scripts to the search path JK> set auto_path [concat $env(HOME)/vmd/scripts/tcl $auto_path] JK> JK> # command extensions (e.g. vmd_draw_vector) have to be sourced JK> foreach ext [glob -nocomplain $env(HOME)/vmd/scripts/extensions/*.tcl ] { JK> source $ext JK> } JK> unset ext JK> ////////////////////////////////////////////////////////////////////// JK> Then made directories named vmd/scripts, vmd/scripts/tcl, JK> vnd/scripts/extensions, and plugins, respectively in My Document directory, JK> and moved addtional files to those directories. for those to work, you have to make sure, that $HOME (or %HOME in DOS notation) is pointing to that 'My Documents' directory. you can also replace the $env(HOME) with any path you like, e.g. "C:\Documents and Settings\user\My Documents\". regards, axel JK> JK> JK> Junichi kato JK> JK> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From utpalchemiitkgp at yahoo.com Mon Apr 4 19:32:19 2005 From: utpalchemiitkgp at yahoo.com (Utpal Sarkar) Date: Mon, 4 Apr 2005 10:32:19 -0700 (PDT) Subject: [CPMD-list] Request for input file Message-ID: <20050404173219.23813.qmail@web51401.mail.yahoo.com> Dear CPMD users, I am begginer of the CPMD code. I have installed it. I also have the manual but I could not understand how to run this code. Or in brief I if anybody help me providing a prototype example of the input file, it will be very helpful for me. I, say, want to see the excited states of HCHO molecule using DFT methods. So how can I proceed? Then how will I define the input files? Any type of help is highly appreciated. Regards Utpal --------------------------------- Do you Yahoo!? Make Yahoo! your home page -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050404/035010e4/attachment.html From dogbe at unr.nevada.edu Mon Apr 4 22:44:51 2005 From: dogbe at unr.nevada.edu (John Kofi Dogbe) Date: Mon, 4 Apr 2005 13:44:51 -0700 (PDT) Subject: [CPMD-list] undefined references to dcopy_ In-Reply-To: References: Message-ID: Greetings from "the biggest little city ...". Please, can anybody help me with the following compilation error/problem? I'm using the evaluation version of the Portland Group Compiler. I'm considering purchasing it for Athlon AMD64 based cluster running RHEL 3.0. In any case, I have compiled CPMD with IFC and the idea is to test the executables from both compilers ... Anyways, I'll appreciate your suggestions on what I'm doing wrong or what I'm missing. I'll also appreciate your experiences on a choice of compiler that can best work for my arch. >>> find below part of the console ouput from make ... mdfile.o(.text+0x1a3f): In function `mdfile_': : undefined reference to `dcopy_' mdfile.o(.text+0x1a97): more undefined references to `dcopy_' follow kdp_diag.o(.text+0x9a3): In function `kdp_diag_': : undefined reference to `zhpev_' kdp_rho.o(.text+0x8df): In function `kdp_rho_': : undefined reference to `zgemm_' kdpoints.o(.text+0x23c): In function `kdpoints_': : undefined reference to `dcopy_' kdpoints.o(.text+0x25f): In function `kdpoints_': : undefined reference to `dcopy_' mm_cpmd_esp_charges_f77.o(.text+0x2e8): In function `mm_cpmd_esp_charges_f77_': : undefined reference to `dcopy_' bs_forces_diag.o(.text+0xd5e): In function `bs_forces_diag_': : undefined reference to `dcopy_' bswfo.o(.text+0x28a): In function `bs_wfo_': : undefined reference to `dcopy_' bswfo.o(.text+0x79f): more undefined references to `dcopy_' follow make: *** [cpmd.x] Error 2 Thanks in advance. John -- **The first is not necessarily the Leader** "If I have spoken evil, bear witness of the evil: but if well, why smitest thou me?" -- Jesus Christ (John, 18:23) From jiten at postech.ac.kr Tue Apr 5 08:14:05 2005 From: jiten at postech.ac.kr (Jiten) Date: Tue, 5 Apr 2005 15:14:05 +0900 Subject: [CPMD-list] Neutralize the excess charge in CPMD Message-ID: <10fc01c539a6$ab69d000$447fdf8d@theo68> Dear CPMD users, In few limited papers related to CPMD for excess charged systems, it hardly explain about the use of neutalizing the excess charge, a uniform background opposite charge are added to the system. I look at the CPMD manual as well, there is no section describing this option. I am just wondering whenever we do CPMD MD for the charged systems using box type of HOCKNEY Method or TUCKERMAN Method, are we taken cared to neutralize the excess charge. Could anyone suggest a reference and how can I deal with CPMD related to this. Sincerely, N. Jiten Singh C/O Prof. Kwang S. Kim Department of Chemistry Pohang University of Science and Technology San 31, Hyojadong, Namgu Pohang 790-784, Korea Phone : 82-54-279-5853 ( Lab ) / 279-4138 ( Appt ) Fax : 82-54-279-8137 (or +82-54-279-3399) Web : http://csm.postech.ac.kr/ and http://www.postech.ac.kr/e Home Page : http://www.geocities.com/njs_19 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050405/8020c97d/attachment.html From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Tue Apr 5 08:39:54 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Tue, 5 Apr 2005 08:39:54 +0200 (CEST) Subject: [CPMD-list] Re: undefined references to dcopy_ In-Reply-To: Message-ID: On Mon, 4 Apr 2005, John Kofi Dogbe wrote: JD> Greetings from "the biggest little city ...". hi john, JD> Please, can anybody help me with the following compilation error/problem? JD> I'm using the evaluation version of the Portland Group Compiler. I'm JD> considering purchasing it for Athlon AMD64 based cluster running RHEL 3.0. JD> In any case, I have compiled CPMD with IFC and the idea is to test the JD> executables from both compilers ... in my tests, the EM64T version of the intel compiler was on average about 5% faster. the PGI compilers have some problems with miscompiling code occasionally, but that happens mostly when using 'new' fortran90 features. JD> Anyways, I'll appreciate your suggestions on what I'm doing wrong or what JD> I'm missing. I'll also appreciate your experiences on a choice of compiler JD> that can best work for my arch. JD> JD> >>> find below part of the console ouput from make ... JD> JD> mdfile.o(.text+0x1a3f): In function `mdfile_': JD> : undefined reference to `dcopy_' JD> mdfile.o(.text+0x1a97): more undefined references to `dcopy_' follow JD> kdp_diag.o(.text+0x9a3): In function `kdp_diag_': JD> : undefined reference to `zhpev_' ... those are all BLAS/LAPACK subroutines. you need to link to a library that provides those functions, e.g. ATLAS, ACML, or (with the least preference BLAS/LAPACK). with the intel compilers, you can also use the EM64T version of the MKL (i have successfully tested version 7.2.1pl3), it is roughly of the same speed as ATLAS and ACML. [...] regards, axel. JD> make: *** [cpmd.x] Error 2 JD> JD> JD> Thanks in advance. JD> John JD> -- JD> **The first is not necessarily the Leader** JD> JD> "If I have spoken evil, bear witness of the evil: JD> but if well, why smitest thou me?" JD> -- Jesus Christ (John, 18:23) JD> JD> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From galia_mex at yahoo.com Tue Apr 5 11:42:30 2005 From: galia_mex at yahoo.com (Galina K.) Date: Tue, 5 Apr 2005 04:42:30 -0500 (CDT) Subject: [CPMD-list] LDA potential Message-ID: <20050405094230.57131.qmail@web53607.mail.yahoo.com> Dear all! Can someone recommend a LDA pseudopotential for Ni, different from Ni_MH? Thanx! Galina _________________________________________________________ Do You Yahoo!? Informaci?n de Estados Unidos y Am?rica Latina, en Yahoo! Noticias. Vis?tanos en http://noticias.espanol.yahoo.com From knba713 at hotmail.com Tue Apr 5 20:04:33 2005 From: knba713 at hotmail.com (Xudong Tang) Date: Tue, 05 Apr 2005 14:04:33 -0400 Subject: [CPMD-list] VMD electronic density contour Message-ID: Hello - I used IsoSurface & VolumeSlice in VMD to show the electronic density around some moleculars, such as H2O. It uses color difference to show different electronic density. However, I'm wondering if I can establish a legend aside, because it's really hard to tell the value of electronic density each color represents. Thanks. Xudong From irg20 at chem.ubc.ca Mon Apr 11 18:56:12 2005 From: irg20 at chem.ubc.ca (Ian R. Greig) Date: Mon, 11 Apr 2005 09:56:12 -0700 (PDT) Subject: [CPMD-list] Re: Equilibration of organic molecules in water In-Reply-To: <200503041100.MAA29908@cpmd.org> Message-ID: Dear all, I was wondering whether anyone has any further comments on the various methods for equilibrating solute molecules in water. The methods that seem to have been used are... 1) Perform a classical MD equilibration of the system prior to any CPMD simulation as suggested by Axel. This equilibration I assume is carried out in the XXX ensemble and the quantum box size used is the same as that from the classical equilibration. 2) Equilibrate a quantum ------------------------------------------------------------------ Ian R. Greig Chemistry Department Room 209, Green College Room E257 2036 Main Mall 6201 Cecil Green Park Road Vancouver, British Columbia Vancouver, British Columbia CANADA V6T 1Z1 CANADA V6T 1Z1 Tel: + 1 604 822 4626 Tel: + 1 604 221 1971 Fax: + 1 604 822 2847 irg20 at cantab.net irg20 at chem.ubc.ca ------------------------------------------------------------------ From kallies at zib.de Tue Apr 12 20:41:19 2005 From: kallies at zib.de (Bernd Kallies) Date: Tue, 12 Apr 2005 20:41:19 +0200 Subject: [CPMD-list] MPI-Problem with WF optimization Message-ID: <1113331278.4960.355.camel@kallies.zib.de> Dear all, I ran into a curious problem when running the attached input with CPMD v3.9.1 (downloaded 3. June 2004) on an IBM p690. The run aborts in a wavefunction optimization with the error msg from the MPI-library: ERROR: 0032-117 User pack or receive buffer is too small (32768) in MPI_Allreduce, task 0 The error occurs in different stages, depending on the number of tasks or machine state. Buffer sizes that are detected to be wrong for MPI_Allreduce differ. Debugging showed a code problem, which seems to be fundamental to me. The error is generated because MPI_Allreduce is called by MPI tasks that are out of sync (calling glosum in different code contexts). The master task is in different context than the others. The reason for that is that the variable TNOFOR (set in tol_chk_cnvener) evaluates to different results (task 0 different from other tasks). The reason for that is that task 0 has a different total energy than the others on entry of tol_chk_cnvener. And the reason for that is, that subroutine linesr (pcgrad.f) contains the lines IF(PARENT) THEN CALL EBACK(0) ENDIF ... IF(PARENT) THEN CALL EBACK(1) ENDIF This yields different total energies for the MPI tasks when checking for wavefunction convergence. When letting all tasks backing up and restoring energy values in linesr, the error mentioned disappears, and the calculation finishes properly. It is not really clear to me which impact this finding has, since line-searching wavefunctions is a task that is done in many cpmd runs. --the input-- &CPMD OPTIMIZE GEOMETRY LBFGS PCG MINIMIZE CONVERGENCE ORBITALS 1.d-7 CONVERGENCE ADAPT 0.02 CONVERGENCE ENERGY 0.05 STORE 50 TASKGROUPS 1 &END &DFT FUNCTIONAL PBE GC-CUTOFF 1.d-6 &END &SYSTEM ANGSTROM SYMMETRY 8 CELL ABSOLUTE 8.34 8.34 14.255 0 0 0 CUTOFF 40.0 TESR 4 DUAL 6.0 &END &ATOMS *Mg_VDB_PBE.psp BINARY NEWF LMAX=D 32 0.000 0.000 0.000 4.170 0.000 0.000 2.085 2.085 0.000 [bzfbbk at berni1 test]> cat inp-geo &CPMD OPTIMIZE GEOMETRY LBFGS PCG MINIMIZE CONVERGENCE ORBITALS 1.d-7 CONVERGENCE ADAPT 0.02 CONVERGENCE ENERGY 0.05 STORE 50 TASKGROUPS 1 &END &DFT FUNCTIONAL PBE GC-CUTOFF 1.d-6 &END &SYSTEM ANGSTROM SYMMETRY 8 CELL ABSOLUTE 8.34 8.34 14.255 0 0 0 CUTOFF 40.0 TESR 4 DUAL 6.0 &END &ATOMS *Mg_VDB_PBE.psp BINARY NEWF LMAX=D 32 0.000 0.000 0.000 4.170 0.000 0.000 2.085 2.085 0.000 6.255 2.085 0.000 0.000 4.170 0.000 4.170 4.170 0.000 2.085 6.255 0.000 6.255 6.255 0.000 2.085 0.000 2.085 6.255 0.000 2.085 0.000 2.085 2.085 4.170 2.085 2.085 2.085 4.170 2.085 6.255 4.170 2.085 0.000 6.255 2.085 4.170 6.255 2.085 0.000 0.000 4.170 4.170 0.000 4.170 2.085 2.085 4.170 6.255 2.085 4.170 0.000 4.170 4.170 4.170 4.170 4.170 2.085 6.255 4.170 6.255 6.255 4.170 2.085 0.000 6.255 6.255 0.000 6.255 0.000 2.085 6.255 4.170 2.085 6.255 2.085 4.170 6.255 6.255 4.170 6.255 0.000 6.255 6.255 4.170 6.255 6.255 *O_VDB_PBE.psp BINARY NEWF LMAX=D 32 2.085 0.000 0.000 6.255 0.000 0.000 0.000 2.085 0.000 4.170 2.085 0.000 2.085 4.170 0.000 6.255 4.170 0.000 0.000 6.255 0.000 4.170 6.255 0.000 0.000 0.000 2.085 4.170 0.000 2.085 2.085 2.085 2.085 6.255 2.085 2.085 0.000 4.170 2.085 4.170 4.170 2.085 2.085 6.255 2.085 6.255 6.255 2.085 2.085 0.000 4.170 6.255 0.000 4.170 0.000 2.085 4.170 4.170 2.085 4.170 2.085 4.170 4.170 6.255 4.170 4.170 0.000 6.255 4.170 4.170 6.255 4.170 0.000 0.000 6.255 4.170 0.000 6.255 2.085 2.085 6.255 6.255 2.085 6.255 0.000 4.170 6.255 4.170 4.170 6.255 2.085 6.255 6.255 6.255 6.255 6.255 CONSTRAINTS FIX ATOMES 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 END CONSTRAINTS &END -- Dr. Bernd Kallies Konrad-Zuse-Zentrum f?r Informationstechnik Berlin Takustr. 7 14195 Berlin Tel: +49-30-84185-270 Fax: +49-30-84185-311 e-mail: kallies at zib.de From chongsin at faculty.nctu.edu.tw Wed Apr 13 17:30:40 2005 From: chongsin at faculty.nctu.edu.tw (Chongsin_Faculty) Date: Wed, 13 Apr 2005 23:30:40 +0800 Subject: [CPMD-list] soliciting a lecturer on CPMD in summer 2005 Message-ID: <003001c5403d$c0a049c0$268ffea9@musta8wda522ki> Dear All: We are running a First Principle Group (FPG) in Taiwan, which consists of physicist, mathematician and engineering faculties and students. We are interested in learning the first-hand details of CPMD code from experts in this field and possible future cooperation in this regard. We can provide flight, accomodation and fair amount of honorarium for your stay. Expected stay period would be 1 week or so in the summer 2005. If you are interested in this or you think someone you know of maybe interested in this, pls kindly contact me via e-mail. Thanks for your help in advance. Sincerely yours, Chongsin ================================================= J.-S. Wu (aka. Chongsin Gou), Assoc. Professor Department of Mechanical Engineering National Chiao-Tung University Hsinchu 30050, TAIWAN Voice: +886-3-573-1693 Fax: +886-3-572-0634 Cell: +886-920-978-043 E-mail: chongsin at faculty.nctu.edu.tw URL: http://wwwdata.me.nctu.edu.tw/must/ ================================================== -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050413/7962e0f9/attachment.html From cur at zurich.ibm.com Thu Apr 14 13:06:44 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Thu, 14 Apr 2005 13:06:44 +0200 Subject: [CPMD-list] MPI-Problem with WF optimization In-Reply-To: <1113331278.4960.355.camel@kallies.zib.de> Message-ID: Bernd, thank you for input - this is a known bug and has been corrected in the soon to be next minor release. REgards Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 Bernd Kallies Sent by: cpmd-list-bounces at cpmd.org 04/12/2005 08:41 PM To cpmd-list at cpmd.org cc Subject [CPMD-list] MPI-Problem with WF optimization Dear all, I ran into a curious problem when running the attached input with CPMD v3.9.1 (downloaded 3. June 2004) on an IBM p690. The run aborts in a wavefunction optimization with the error msg from the MPI-library: ERROR: 0032-117 User pack or receive buffer is too small (32768) in MPI_Allreduce, task 0 The error occurs in different stages, depending on the number of tasks or machine state. Buffer sizes that are detected to be wrong for MPI_Allreduce differ. Debugging showed a code problem, which seems to be fundamental to me. The error is generated because MPI_Allreduce is called by MPI tasks that are out of sync (calling glosum in different code contexts). The master task is in different context than the others. The reason for that is that the variable TNOFOR (set in tol_chk_cnvener) evaluates to different results (task 0 different from other tasks). The reason for that is that task 0 has a different total energy than the others on entry of tol_chk_cnvener. And the reason for that is, that subroutine linesr (pcgrad.f) contains the lines IF(PARENT) THEN CALL EBACK(0) ENDIF ... IF(PARENT) THEN CALL EBACK(1) ENDIF This yields different total energies for the MPI tasks when checking for wavefunction convergence. When letting all tasks backing up and restoring energy values in linesr, the error mentioned disappears, and the calculation finishes properly. It is not really clear to me which impact this finding has, since line-searching wavefunctions is a task that is done in many cpmd runs. --the input-- &CPMD OPTIMIZE GEOMETRY LBFGS PCG MINIMIZE CONVERGENCE ORBITALS 1.d-7 CONVERGENCE ADAPT 0.02 CONVERGENCE ENERGY 0.05 STORE 50 TASKGROUPS 1 &END &DFT FUNCTIONAL PBE GC-CUTOFF 1.d-6 &END &SYSTEM ANGSTROM SYMMETRY 8 CELL ABSOLUTE 8.34 8.34 14.255 0 0 0 CUTOFF 40.0 TESR 4 DUAL 6.0 &END &ATOMS *Mg_VDB_PBE.psp BINARY NEWF LMAX=D 32 0.000 0.000 0.000 4.170 0.000 0.000 2.085 2.085 0.000 [bzfbbk at berni1 test]> cat inp-geo &CPMD OPTIMIZE GEOMETRY LBFGS PCG MINIMIZE CONVERGENCE ORBITALS 1.d-7 CONVERGENCE ADAPT 0.02 CONVERGENCE ENERGY 0.05 STORE 50 TASKGROUPS 1 &END &DFT FUNCTIONAL PBE GC-CUTOFF 1.d-6 &END &SYSTEM ANGSTROM SYMMETRY 8 CELL ABSOLUTE 8.34 8.34 14.255 0 0 0 CUTOFF 40.0 TESR 4 DUAL 6.0 &END &ATOMS *Mg_VDB_PBE.psp BINARY NEWF LMAX=D 32 0.000 0.000 0.000 4.170 0.000 0.000 2.085 2.085 0.000 6.255 2.085 0.000 0.000 4.170 0.000 4.170 4.170 0.000 2.085 6.255 0.000 6.255 6.255 0.000 2.085 0.000 2.085 6.255 0.000 2.085 0.000 2.085 2.085 4.170 2.085 2.085 2.085 4.170 2.085 6.255 4.170 2.085 0.000 6.255 2.085 4.170 6.255 2.085 0.000 0.000 4.170 4.170 0.000 4.170 2.085 2.085 4.170 6.255 2.085 4.170 0.000 4.170 4.170 4.170 4.170 4.170 2.085 6.255 4.170 6.255 6.255 4.170 2.085 0.000 6.255 6.255 0.000 6.255 0.000 2.085 6.255 4.170 2.085 6.255 2.085 4.170 6.255 6.255 4.170 6.255 0.000 6.255 6.255 4.170 6.255 6.255 *O_VDB_PBE.psp BINARY NEWF LMAX=D 32 2.085 0.000 0.000 6.255 0.000 0.000 0.000 2.085 0.000 4.170 2.085 0.000 2.085 4.170 0.000 6.255 4.170 0.000 0.000 6.255 0.000 4.170 6.255 0.000 0.000 0.000 2.085 4.170 0.000 2.085 2.085 2.085 2.085 6.255 2.085 2.085 0.000 4.170 2.085 4.170 4.170 2.085 2.085 6.255 2.085 6.255 6.255 2.085 2.085 0.000 4.170 6.255 0.000 4.170 0.000 2.085 4.170 4.170 2.085 4.170 2.085 4.170 4.170 6.255 4.170 4.170 0.000 6.255 4.170 4.170 6.255 4.170 0.000 0.000 6.255 4.170 0.000 6.255 2.085 2.085 6.255 6.255 2.085 6.255 0.000 4.170 6.255 4.170 4.170 6.255 2.085 6.255 6.255 6.255 6.255 6.255 CONSTRAINTS FIX ATOMES 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 END CONSTRAINTS &END -- Dr. Bernd Kallies Konrad-Zuse-Zentrum f?r Informationstechnik Berlin Takustr. 7 14195 Berlin Tel: +49-30-84185-270 Fax: +49-30-84185-311 e-mail: kallies at zib.de _______________________________________________ CPMD-list mailing list CPMD-list at cpmd.org http://cpmd.org/mailman/listinfo/cpmd-list -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050414/f732f2aa/attachment.html From i.kozin at dl.ac.uk Thu Apr 14 14:21:25 2005 From: i.kozin at dl.ac.uk (Kozin, I (Igor)) Date: Thu, 14 Apr 2005 13:21:25 +0100 Subject: [CPMD-list] undefined references to dcopy_ Message-ID: <77673C9ECE12AB4791B5AC0A7BF40C8F1543F6@exchange02.fed.cclrc.ac.uk> Hi John, I totally agree with Axel about the source of your errors. You should certainly be able to get Intel build working. I am sure I've looked at the post Sept 2004, Axel Kohlmeyer: Opteron + Intel Compiler EM64T My experience with Intel compilers 8.1 for EM64T and CPMD on AMD64 was very good. Another compiler worth considering is PathScale. However I must admit that at the time I did my benchmarking Intel 8.1 outperformed PathScale 1.4 on wat32 benchmark (which in turn outperformed PGI 5.2). However, since that all three released major updates: Intel 9.0, PathScale 2.0 and PGI 6.0. Make sure you are using the latest versions of each. Best, ink I. Kozin (i.kozin at dl.ac.uk) CCLRC Daresbury Laboratory tel: 01925 603308 http://www.cse.clrc.ac.uk/disco > -----Original Message----- > From: cpmd-list-bounces at cpmd.org [mailto:cpmd-list-bounces at cpmd.org]On > Behalf Of John Kofi Dogbe > Sent: 04 April 2005 21:45 > To: Axel Kohlmeyer > Cc: cpmd-list at cpmd.org > Subject: [CPMD-list] undefined references to dcopy_ > > > Greetings from "the biggest little city ...". > > Please, can anybody help me with the following compilation > error/problem? > I'm using the evaluation version of the Portland Group Compiler. I'm > considering purchasing it for Athlon AMD64 based cluster > running RHEL 3.0. > In any case, I have compiled CPMD with IFC and the idea is to test the > executables from both compilers ... > > Anyways, I'll appreciate your suggestions on what I'm doing > wrong or what > I'm missing. I'll also appreciate your experiences on a > choice of compiler > that can best work for my arch. > > >>> find below part of the console ouput from make ... > > mdfile.o(.text+0x1a3f): In function `mdfile_': > : undefined reference to `dcopy_' > mdfile.o(.text+0x1a97): more undefined references to `dcopy_' follow > kdp_diag.o(.text+0x9a3): In function `kdp_diag_': > : undefined reference to `zhpev_' > kdp_rho.o(.text+0x8df): In function `kdp_rho_': > : undefined reference to `zgemm_' > kdpoints.o(.text+0x23c): In function `kdpoints_': > : undefined reference to `dcopy_' > kdpoints.o(.text+0x25f): In function `kdpoints_': > : undefined reference to `dcopy_' > mm_cpmd_esp_charges_f77.o(.text+0x2e8): In function > `mm_cpmd_esp_charges_f77_': > : undefined reference to `dcopy_' > bs_forces_diag.o(.text+0xd5e): In function `bs_forces_diag_': > : undefined reference to `dcopy_' > bswfo.o(.text+0x28a): In function `bs_wfo_': > : undefined reference to `dcopy_' > bswfo.o(.text+0x79f): more undefined references to `dcopy_' follow > make: *** [cpmd.x] Error 2 > > > Thanks in advance. > John > -- > **The first is not necessarily the Leader** > > "If I have spoken evil, bear witness of the evil: > but if well, why smitest thou me?" > -- Jesus Christ (John, 18:23) > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list > From hutter at pci.unizh.ch Thu Apr 14 14:41:33 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Thu, 14 Apr 2005 14:41:33 +0200 (MEST) Subject: [CPMD-list] Request for input file In-Reply-To: <20050404173219.23813.qmail@web51401.mail.yahoo.com> References: <20050404173219.23813.qmail@web51401.mail.yahoo.com> Message-ID: Hi there are many input and output files available on the cpmd.org webpage. Go to the download area and there to the contrib folder. The cpmd-test files has the inputs. regards Juerg Hutter ---------------------------------------------------------- Juerg Hutter Phone : ++41 1 635 4491 Physical Chemistry Institute FAX : ++41 1 635 6838 University of Zurich E-mail: hutter at pci.unizh.ch Winterthurerstrasse 190 CH-8057 Zurich, Switzerland ---------------------------------------------------------- On Mon, 4 Apr 2005, Utpal Sarkar wrote: > Dear CPMD users, > I am begginer of the CPMD code. I have installed it. I also have > the manual but I could not understand how to run this code. Or in brief I if anybody > help me providing a prototype example of the input file, it will be very helpful for me. > I, say, want to see the excited states of HCHO molecule using DFT methods. > So how can I proceed? Then how will I define the input files? > Any type of help is highly appreciated. > Regards > Utpal > > > --------------------------------- > Do you Yahoo!? > Make Yahoo! your home page From hutter at pci.unizh.ch Thu Apr 14 14:50:08 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Thu, 14 Apr 2005 14:50:08 +0200 (MEST) Subject: [CPMD-list] Neutralize the excess charge in CPMD In-Reply-To: <10fc01c539a6$ab69d000$447fdf8d@theo68> References: <10fc01c539a6$ab69d000$447fdf8d@theo68> Message-ID: Hi if there is a net charge in your system CPMD takes care of this depending on the Poisson solver used. For periodic systems a neutralizing background charge is added (no keyword option is needed, this is done automatically). If you use a Poisson solver for isolated systems (Hockney or Tuckerman) then the background charge is not needed. These methods calculate correctly the G=0 term of the potential. This is explained in the original papers or in the review at D. Marx and J. Hutter published in Modern Methods and Algorithms of Quantum Chemistry, Forschungszentrum J?lich, NIC Series, Vol. 1 (2000), 301-449 http://www.fz-juelich.de/nic-series/Volume1/ regards Juerg Hutter ---------------------------------------------------------- Juerg Hutter Phone : ++41 1 635 4491 Physical Chemistry Institute FAX : ++41 1 635 6838 University of Zurich E-mail: hutter at pci.unizh.ch Winterthurerstrasse 190 CH-8057 Zurich, Switzerland ---------------------------------------------------------- On Tue, 5 Apr 2005, Jiten wrote: > Dear CPMD users, > > In few limited papers related to CPMD for excess charged systems, it hardly explain about the use of neutalizing the excess charge, a uniform background opposite charge are added to the system. > > I look at the CPMD manual as well, there is no section describing this option. > > I am just wondering whenever we do CPMD MD for the charged systems using box type of HOCKNEY Method or TUCKERMAN Method, are we taken cared to neutralize the excess charge. > > Could anyone suggest a reference and how can I deal with CPMD related to this. > > Sincerely, > > N. Jiten Singh > C/O Prof. Kwang S. Kim > Department of Chemistry > Pohang University of Science and Technology > San 31, Hyojadong, Namgu > Pohang 790-784, Korea > Phone : 82-54-279-5853 ( Lab ) / 279-4138 ( Appt ) > Fax : 82-54-279-8137 (or +82-54-279-3399) > Web : http://csm.postech.ac.kr/ and http://www.postech.ac.kr/e > Home Page : http://www.geocities.com/njs_19 From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Thu Apr 14 15:01:53 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Thu, 14 Apr 2005 15:01:53 +0200 (CEST) Subject: [CPMD-list] undefined references to dcopy_ In-Reply-To: <77673C9ECE12AB4791B5AC0A7BF40C8F1543F6@exchange02.fed.cclrc.ac.uk> Message-ID: On Thu, 14 Apr 2005, Kozin, I (Igor) wrote: hi igor, ...another 'trick' worth trying to get the last bit of performance out of opteron machines is to use an ATLAS library (where you can gain a little on the side by enlarging the internal buffers) in combination with the math library from the ACML (i.e. only use -lacml_mv) [...] IK> IK> However, since that all three released major updates: IK> Intel 9.0, PathScale 2.0 and PGI 6.0. AFAIK, intel 9.0 is still in beta. IK> Make sure you are using the latest versions of each. regards, axel. -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Thu Apr 14 15:09:41 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Thu, 14 Apr 2005 15:09:41 +0200 (CEST) Subject: [CPMD-list] Request for input file In-Reply-To: <20050404173219.23813.qmail@web51401.mail.yahoo.com> Message-ID: On Mon, 4 Apr 2005, Utpal Sarkar wrote: US> Dear CPMD users, dear utpal, US> I am begginer of the CPMD code. I have US> installed it. I also have the manual but I could not understand how to US> run this code. Or in brief I if anybody help me providing a prototype US> example of the input file, it will be very helpful for me. I, say, there are some examples in the CPMD-test archive at cpmd.org and there is a proton transport in water example (adapted from our local theory lab course) and a very basic (work-in-progress) tutorial available from my personal homepage. US> want to see the excited states of HCHO molecule using DFT methods. So US> how can I proceed? Then how will I define the input files? Any type of US> help is highly appreciated. Regards Utpal frankly, these kind of calculations are can easily become quite tricky. so before tackling that, you should first become familiar with the 'normal' types of CPMD calculations and take your time in learning them. regards, axel. US> US> US> --------------------------------- US> Do you Yahoo!? US> Make Yahoo! your home page -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Thu Apr 14 15:19:05 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Thu, 14 Apr 2005 15:19:05 +0200 (CEST) Subject: [CPMD-list] VMD electronic density contour In-Reply-To: Message-ID: On Tue, 5 Apr 2005, Xudong Tang wrote: XT> Hello - dear xudong, this is a VMD related question and you should better ask it on the VMD mailing list. XT> I used IsoSurface & VolumeSlice in VMD to show the electronic density XT> around some moleculars, such as H2O. It uses color difference to show XT> different electronic density. However, I'm wondering if I can XT> establish a legend aside, because it's really hard to tell the value XT> of electronic density each color represents. that kind of legend has been asked for several times, but it is not (yet) implemented and, to be frank, for the way the output is produced in VMD it is of limited use. if you want to map out the colors more exactly, you should turn to a tool like OpenDX, AVS or alike. if you want to stick to VMD, you could create multiple isosurfaces at specific isovalues, use a different color for each of them and then have a clip plane cutting through them. with VMD 1.8.3 you have a clipping plane tool plugin from the extension menu to do that. but for that kind of visualization, pymol is probably better suited (i've seen pictures). regards, axel. XT> XT> Thanks. XT> XT> Xudong XT> XT> XT> _______________________________________________ XT> CPMD-list mailing list XT> CPMD-list at cpmd.org XT> http://cpmd.org/mailman/listinfo/cpmd-list XT> XT> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Thu Apr 14 15:28:26 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Thu, 14 Apr 2005 15:28:26 +0200 (CEST) Subject: [CPMD-list] LDA potential In-Reply-To: <20050405094230.57131.qmail@web53607.mail.yahoo.com> Message-ID: On Tue, 5 Apr 2005, Galina K. wrote: GK> Dear all! dear galina, GK> Can someone recommend a LDA pseudopotential for Ni, GK> different from Ni_MH? i am not aware of a Ni_MH potential file, in any of the 'standard' cpmd pseudopotential libraries, where did you get it? and what kind of problem(s) do you have (if any)? to recommend a different one, it would be useful to know what kind of calculation of what kind of system you want to do, i.e. whether you need a semi-core potential or not. Ni is 'tricky' anyways. if you don't find what you need at cpmd.org, you may want to check out the information page on juerg hutters homepage: http://www.pci.unizh.ch/gruppe.hutter/e/information/ which has additional, recently updated pseudopotential libraries. regards, axel. GK> Thanx! GK> Galina GK> GK> _________________________________________________________ GK> Do You Yahoo!? GK> Informaci?n de Estados Unidos y Am?rica Latina, en Yahoo! Noticias. GK> Vis?tanos en http://noticias.espanol.yahoo.com GK> _______________________________________________ GK> CPMD-list mailing list GK> CPMD-list at cpmd.org GK> http://cpmd.org/mailman/listinfo/cpmd-list GK> GK> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From irg20 at chem.ubc.ca Thu Apr 14 20:56:00 2005 From: irg20 at chem.ubc.ca (Ian R. Greig) Date: Thu, 14 Apr 2005 11:56:00 -0700 (PDT) Subject: [CPMD-list] Re: Equilibration of organic molecules in water Message-ID: Dear all, Apologies for the partially formed message sent previously. I was wondering whether anyone has any further comments on the various methods for equilibrating solute molecules in water. The methods that seem to have been used are... 1) Perform a classical MD equilibration of the system prior to any CPMD simulation as suggested by Axel. This equilibration I assume is carried out in the NPT ensemble and the quantum box size used is the same as that from the classical equilibration. 2) Equilibrate a quantum system (presumably either derived as in 1 above or by replacement of some number of water atoms but the solute of interest) using Parrinello-Rahman dynamics (e.g. as per Gaigeot and Sprik - J. Phys. Chem. B 107, 10344, 203) to obtain the appropriate box size. Does anyone have comments on the relative suitability and computational expense of both methods? I imagine that method 2) (probably starting from an MM-derived system) is probably the most rigorous but that 1) will probably be alot quicker. Additionally, it seems that any study of the reactivity of a small system will probably be practically limited to Car-Parrinello dynamics. This will ignore any change in volume on reaction which might be quite significant? Many thanks for your help and comments as always, Ian ------------------------------------------------------------------ Ian R. Greig Chemistry Department Room E257 2036 Main Mall Vancouver, British Columbia CANADA V6T 1Z1 Tel: + 1 604 822 4626 Fax: + 1 604 822 2847 irg20 at cantab.net irg20 at chem.ubc.ca ------------------------------------------------------------------ From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Thu Apr 14 22:12:16 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Thu, 14 Apr 2005 22:12:16 +0200 Subject: [CPMD-list] Re: Equilibration of organic molecules in water In-Reply-To: Your message of "Thu, 14 Apr 2005 11:56:00 PDT." Message-ID: <200504142012.j3EKCG814316@yello.theochem.ruhr-uni-bochum.de> >>> "IG" == Ian R Greig writes: IG> Dear all, dear ian, IG> I was wondering whether anyone has any further comments on the various IG> methods for equilibrating solute molecules in water. IG> The methods that seem to have been used are... IG> 1) Perform a classical MD equilibration of the system prior to any CPMD IG> simulation as suggested by Axel. This equilibration I assume is carried IG> out in the NPT ensemble and the quantum box size used is the same as that IG> from the classical equilibration. yes. of course this can only work if you have adequate force field parameters for the molecule in question. in my tests using SPC/E as water potential seemed to produce a water structure, that matches the one of BLYP water quite closely. IG> 2) Equilibrate a quantum system (presumably either derived as in 1 above IG> or by replacement of some number of water atoms but the solute of IG> interest) using Parrinello-Rahman dynamics (e.g. as per Gaigeot and Sprik IG> - J. Phys. Chem. B 107, 10344, 203) to obtain the appropriate box size. what has been done, is to take a well equilibrated water restart, determine the size of a hole needed for the molecule in question and find the closest matching water cluster to be removed. of course this works best for small molecules. the most tricky setup to equilibrate with CPMD, that i have encountered so far, was a dense water layer inbetween a crystal slab. IG> Does anyone have comments on the relative suitability and computational IG> expense of both methods? I imagine that method 2) (probably starting from IG> an MM-derived system) is probably the most rigorous but that 1) will IG> probably be alot quicker. doing a parrinello-rahman run with CPMD is _very_ expensive. you need to make sure, that the stress tensor is well converged with respect to the plane wave cutoff, and that takes very much larger plane wave cutoff than for a normal CP-MD. but also the run is then done with a constant number of plane waves, which means that the effective cutoff changes with the size of the box. in short, you can do it, but you want to avoid it unless you have no other way. IG> Additionally, it seems that any study of the reactivity of a small system IG> will probably be practically limited to Car-Parrinello dynamics. This will IG> ignore any change in volume on reaction which might be quite significant? could you elaborate a little more on this. there are quite a few ways to determine e.g. free energy surfaces, some of which are implemented in cpmd. but of course any reaction that happens spontaneously and without any additional driving in the timeframe available to CPMD simulations, has to be very fast (e.g. proton transfers) and i cannot imagine how something like that would result in a large change in volume. regards, axel. IG> Many thanks for your help and comments as always, IG> Ian IG> ------------------------------------------------------------------ IG> Ian R. Greig IG> Chemistry Department IG> Room E257 2036 Main Mall IG> Vancouver, British Columbia IG> CANADA V6T 1Z1 IG> Tel: + 1 604 822 4626 IG> Fax: + 1 604 822 2847 IG> irg20 at cantab.net IG> irg20 at chem.ubc.ca IG> ------------------------------------------------------------------ IG> _______________________________________________ IG> CPMD-list mailing list IG> CPMD-list at cpmd.org IG> http://cpmd.org/mailman/listinfo/cpmd-list -- ======================================================================= Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From irg20 at chem.ubc.ca Thu Apr 14 23:49:37 2005 From: irg20 at chem.ubc.ca (Ian R. Greig) Date: Thu, 14 Apr 2005 14:49:37 -0700 (PDT) Subject: [CPMD-list] Re: Equilibration of organic molecules in water In-Reply-To: <200504142012.j3EKCG814316@yello.theochem.ruhr-uni-bochum.de> Message-ID: > > IG> Additionally, it seems that any study of the reactivity of a small > system will probably be practically limited to Car-Parrinello > dynamics. This will ignore any change in volume on reaction which > might be quite significant? > > could you elaborate a little more on this. there are quite a few ways > to determine e.g. free energy surfaces, some of which are implemented in > cpmd. but of course any reaction that happens spontaneously and without > any additional driving in the timeframe available to CPMD simulations, > has to be very fast (e.g. proton transfers) and i cannot imagine how > something like that would result in a large change in volume. > Sorry for just throwing this in at the end. I was referring to a system in which something other than a spontaneous proton tranfer or the like was occurring. For example, a bonding change between heavy atom, which would require use of one of the implemented means of driving the system along the reaction coordinate of interest. The point I was trying to make was that such processes (consider the oft-used symmetric displacement reaction occurring at a carbon center in solution for example) might be more likely to involve a significant change in volume at constant pressure along the reaction coordinate given that computationally feasible systems will use a solute surrounded by only a couple of solvation shells. All the best, Ian ------------------------------------------------------------------ Ian R. Greig Chemistry Department Room 209, Green College Room E257 2036 Main Mall 6201 Cecil Green Park Road Vancouver, British Columbia Vancouver, British Columbia CANADA V6T 1Z1 CANADA V6T 1Z1 Tel: + 1 604 822 4626 Tel: + 1 604 221 1971 Fax: + 1 604 822 2847 irg20 at cantab.net irg20 at chem.ubc.ca ------------------------------------------------------------------ From radbalu at comcast.net Fri Apr 15 02:36:19 2005 From: radbalu at comcast.net (Rad Balu) Date: Thu, 14 Apr 2005 20:36:19 -0400 Subject: [CPMD-list] (no subject) Message-ID: <001a01c54153$243cc2e0$0edc0681@D3B9FD51> Dear All, I am wondering whether it is possible to invoke cpmd programmatically to perform a sequence of single point energy calculations. My intent is to perform such calculations to calculcate intensities and anharmonic frequencies. I would like to have any suggestions on how to go about it if it is feasible at all. Thanks in advance Rad -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050414/a95c183f/attachment.html From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Fri Apr 15 08:52:09 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Fri, 15 Apr 2005 08:52:09 +0200 Subject: [CPMD-list] (no subject) In-Reply-To: Your message of "Thu, 14 Apr 2005 20:36:19 EDT." <001a01c54153$243cc2e0$0edc0681@D3B9FD51> Message-ID: <200504150652.j3F6q9E09104@yello.theochem.ruhr-uni-bochum.de> >>> "RB" == Rad Balu writes: RB> Dear All, RB> I am wondering whether it is possible to invoke cpmd programmatically to RB> perform a sequence of single point energy calculations. My intent is to RB> perform such calculations to calculcate intensities and anharmonic RB> frequencies. I would like to have any suggestions on how to go about it RB> if it is feasible at all. dear rad, i see two ways how you can do this: a) you modify the sourcecode of CPMD and write a 'driver' routine to repeatedly call the subroutine wfopts and then just modify any parameters you see fit (within the constraints of what can be changed without initializing everything from scratch). b) you write a script/program to generate a series of input files and then run them one after another (or in parallel, if you have the resources). if you just move some coordinates around a little bit, you can also just modify the GEOMETRY file between runs and do a RESTART WAVEFUNCTION COORDINATES GEOFILE LATEST with the _same_ input file. this way you can re-use the wavefunction of the previous runs and may be able to save some time. if you get something that works, please consider contributing it back to the cpmd distribution, so others can benefit as well. regards, axel. RB> Thanks in advance RB> Rad -- ======================================================================= Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From ZRLSRB at ch.ibm.com Fri Apr 15 09:40:17 2005 From: ZRLSRB at ch.ibm.com (Salomon Billeter) Date: Fri, 15 Apr 2005 09:40:17 +0200 Subject: [CPMD-list] (no subject) Message-ID: Dear Rad, dear Axel, > RB> I am wondering whether it is possible to invoke cpmd programmatically to > RB> perform a sequence of single point energy calculations. My intent is to > RB> perform such calculations to calculcate intensities and anharmonic > RB> frequencies. I would like to have any suggestions on how to go about it > RB> if it is feasible at all. > > dear rad, > > i see two ways how you can do this: > > a) you modify the sourcecode of CPMD and write a 'driver' > routine to repeatedly call the subroutine wfopts and > then just modify any parameters you see fit (within the > constraints of what can be changed without initializing > everything from scratch). a starting point for this already exists, see c) > b) you write a script/program to generate a series of input > files and then run them one after another (or in parallel, > if you have the resources). > if you just move some coordinates around a little bit, you > can also just modify the GEOMETRY file between runs and do a > RESTART WAVEFUNCTION COORDINATES GEOFILE LATEST with the _same_ > input file. this way you can re-use the wavefunction of the > previous runs and may be able to save some time. c) you can use the EGO interface by choosing INTERFACE EGO in the &CPMD section. This interface is very versatile and not limited to QM/MM. You'll find more information in the file egointer.F and in the manual. There is a simple handshake using files, and both the input and the output are exchanged using files. Best regards, Salomon From cur at zurich.ibm.com Fri Apr 15 23:31:47 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Fri, 15 Apr 2005 23:31:47 +0200 Subject: [CPMD-list] CPMD2005 Message-ID: Dear CPMD user, We would like to inform you about the next CPMD workshop (CPMD2005). As the previous workshop of 2003, the site where we will meet is again the Monte Verita conference center Stefano Franscini of the ETH in Ascona, Switzerland. The dates are from Saturday September 3 to Thursday 8. You can find details about the workshop program and registration and on the cpmd web pages at www.cpmd.org. The deadline for registration and to submit a poster is April 30. Due to limitations to the number of partecipants, we cannot garantee a place for all at this stage. Communication about acceptance will follow shortly after the deadline. The 2005 workshop is special. In September that year Michele Parrinello will reach the age of sixty. This event will be celebrated with a two day symposium honouring Prof Parrinello's scientific accomplishments. We expect many of his colleagues and scientific friends to come and give presentations. Symposia of this kind tend to be somewhat exclusive. For this reason we thought it appropriate to embed the Parrinello Symposium in the the CPMD2005 workshop. In this way the participants to the workshop will be able to attend the symposium. It will take two days of the workshop, the 6th and 7th of September, the second day (the Wednesday) coinciding with Michele's Birthday. There is no conference fee, but all participants will be charged for board and accommodation. The costs are CHF 950.- for a place at the Monte Verita center (5 nights, arriving Saturday leaving Thursday, people registering late may have to be lodged at hotels in town) or CHF 400.- when you prefer to arrange accommodation yourself. Both the CHF 950.- and 400.- sums cover all meals, including breakfast at the conference center and the conference dinner. Ideally, we hope to have to sufficient funds to offer financial support to a limited number of junior participants. We will make announcements regarding this possibility when we have a better view of the budget. Yours sincerely Michiel Sprik Wanda Andreoni Matthias Krack -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050415/4c17e865/attachment.html From mclar at elrond.chem.uni.wroc.pl Mon Apr 18 11:58:43 2005 From: mclar at elrond.chem.uni.wroc.pl (Przemyslaw Dopieralski) Date: Mon, 18 Apr 2005 11:58:43 +0200 (CEST) Subject: [CPMD-list] Vizualization path integral dynamics Message-ID: Hello I begin vizualization of path integral dynamic with vmd 1.8.3 liunux ver. (description on Axel Kohlmayer site). I used Axel's script traj2xyz.pl. At this moment OK. But now I don't know what next. I've got new TRAJECTORY with X but vmd doesn't read it - or maybe I did something wrong. My NEW TRAJECTORY with X looks good - I saw examples on Axel's Kohlmayer webpage. What shoud I do now? Best regards, Przemyslaw Dopieralski University of Wroclaw From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Mon Apr 18 12:47:23 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Mon, 18 Apr 2005 12:47:23 +0200 (CEST) Subject: [CPMD-list] Vizualization path integral dynamics In-Reply-To: Message-ID: On Mon, 18 Apr 2005, Przemyslaw Dopieralski wrote: PD> PD> Hello dear przemyslaw, PD> I begin vizualization of path integral dynamic with vmd 1.8.3 liunux ver. PD> (description on Axel Kohlmayer site). I used Axel's script traj2xyz.pl. PD> At this moment OK. But now I don't know what next. I've got new TRAJECTORY PD> with X but vmd doesn't read it - or maybe I did something wrong. My NEW PD> TRAJECTORY with X looks good - I saw examples on Axel's Kohlmayer webpage. i assume, you are referring to part 6, point 8.1 of the tutorial, i.e.: http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/cpmd-vmd/part6.html#chap8_sect1 the example is just a demontration, that you have to adapt to your specific needs. if you want to get the same or a similar visualization, you can download the nh3-pimd.vmd script, change the line mol new {nh3-pimd.xyz} type xyz waitfor all so it corresponds to the filename of your trajectory and then load the modified vmd script via 'File->Load State...'. you have to realize, that VMD was initially conceived to represent biomolecules and has thus no knowledge or internal heuristics for data from ab initio calculations. so for meaningful visualization of CPMD results, you need to build custom visualizations and add a little additional script magic here and there. to provide some examples and inspiration for that is the main idea of the tutorial. there is _no_ way to have a ready-to-use script for all problems, you have to see it more like a toolkit or a starting point. regards, axel. p.s.: if anybody has suggestions, improvements or visualization problems, that are not covered by the tutorial, please let me know. i'm always looking for new challenges... ;-) PD> PD> What shoud I do now? PD> PD> Best regards, PD> Przemyslaw Dopieralski PD> University of Wroclaw PD> PD> PD> PD> PD> _______________________________________________ PD> CPMD-list mailing list PD> CPMD-list at cpmd.org PD> http://cpmd.org/mailman/listinfo/cpmd-list PD> PD> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From mclar at elrond.chem.uni.wroc.pl Mon Apr 18 15:39:46 2005 From: mclar at elrond.chem.uni.wroc.pl (Przemyslaw Dopieralski) Date: Mon, 18 Apr 2005 15:39:46 +0200 (CEST) Subject: [CPMD-list] Vizualization - again Message-ID: :) hello Still me. AK> the example is just a demontration, that you have to adapt to your AK> specific needs. if you want to get the same or a similar visualization, AK> you can download the nh3-pimd.vmd script, change the line AK> mol new {nh3-pimd.xyz} type xyz waitfor all AK> so it corresponds to the filename of your trajectory and then AK> load the modified vmd script via 'File->Load State...'. OK. I did it many times, but it didn't work. "Unable to load file ...". I changed type, name... but I found that in my new trajectory with X there are additional columns - probably with gradients. In example nh3-pimd.xyz there is only 3 columns and in my new trajectory I've got 6 columns. regards, przemek From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Mon Apr 18 17:40:43 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Mon, 18 Apr 2005 17:40:43 +0200 (CEST) Subject: [CPMD-list] Vizualization - again In-Reply-To: Message-ID: On Mon, 18 Apr 2005, Przemyslaw Dopieralski wrote: PD> :) hello PD> PD> Still me. PD> PD> AK> the example is just a demontration, that you have to adapt to your PD> AK> specific needs. if you want to get the same or a similar PD> visualization, PD> AK> you can download the nh3-pimd.vmd script, change the line PD> PD> AK> mol new {nh3-pimd.xyz} type xyz waitfor all PD> PD> AK> so it corresponds to the filename of your trajectory and then PD> AK> load the modified vmd script via 'File->Load State...'. PD> PD> OK. I did it many times, but it didn't work. PD> "Unable to load file ...". I changed type, name... PD> PD> but I found that in my new trajectory with X there are additional columns PD> - probably with gradients. In example nh3-pimd.xyz there is only 3 columns PD> and in my new trajectory I've got 6 columns. can you please outline the steps you have done to convert the trajectory to .xyz, and/or describe how you load it into VMD and perhaps also post a small sample of the file, so i could check it out myself. regards, axel. PD> PD> regards, PD> przemek PD> PD> _______________________________________________ PD> CPMD-list mailing list PD> CPMD-list at cpmd.org PD> http://cpmd.org/mailman/listinfo/cpmd-list PD> PD> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From alex.merchant at rmit.edu.au Tue Apr 19 08:00:26 2005 From: alex.merchant at rmit.edu.au (Alex Merchant) Date: Tue, 19 Apr 2005 16:00:26 +1000 Subject: [CPMD-list] Compilation on Power5 Message-ID: Hi All, Has anyone gotten CPMD to compile on a Power5 system? Any clues on how to edit the makefile to get it up and running? Alex Merchant == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. From alex.merchant at rmit.edu.au Tue Apr 19 08:34:26 2005 From: alex.merchant at rmit.edu.au (Alex Merchant) Date: Tue, 19 Apr 2005 16:34:26 +1000 Subject: [CPMD-list] Update to previous question Message-ID: Hi Again, I edited the Makefile pertinent to the available compilers (see http://www.vpac.org/content/systems_and_support/support/how_to/edda_start.php for details). I've made a mkefile with the following: #---------------------------------------------------------------------------- # Makefile for cpmd.x (plane wave electronic calculation) # Configuration: IBM-SP4-SMP-64 # Creation of Makefile: Apr 19 2005 # on Linux edda 2.6.5-7.145-pseries64 #1 SMP Thu Jan 27 09:19:29 UTC 2005 ppc64 ppc64 ppc64 GNU/Linux # Author: alex #---------------------------------------------------------------------------- # SHELL = /bin/sh # #--------------- Default Configuration for IBM-SP5-SMP-64 --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 CPP = /usr/local/IBM_compilers/vacpp/7.0/bin/xlc++ -P CPPFLAGS = -D__IBM -DLAPACK -DPOINTER8 -DPARALLEL=PARALLEL -DMP_LIBRARY=__MPI -DFFT_ESSL NOOPT_FLAG =CC = xlc_r FC = xlf_r -c -O3 -qstrict LD = xlf_r -O3 -qstrict AR = /usr/bin/ar #---------------------------------------------------------------------------- CFGDEST = . CFGMACH = IBM-SP5-SMP-64 # # LIST OF FILES [etc] The output I get from the compiler is: xlf_r -c -O3 -qstrict -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp ./cpmd.f -o ./cpmd.o ** cpmd_stuttgart === End of Compilation 1 === "./cpmd.f", line 25.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". "./cpmd.f", line 26.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". ** cpmd === End of Compilation 2 === 1501-511 Compilation failed for file cpmd.f. make: *** [cpmd.o] Error 1 Alex Merchant == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. From cur at zurich.ibm.com Tue Apr 19 12:23:20 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Tue, 19 Apr 2005 12:23:20 +0200 Subject: Fw: [CPMD-list] Update to previous question Message-ID: Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 ----- Forwarded by Alessandro Curioni/Zurich/IBM on 04/19/2005 12:23 PM ----- Alessandro Curioni/Zurich/IBM 04/19/2005 11:56 AM To "Alex Merchant" cc Subject Re: [CPMD-list] Update to previous question Alex, in your Makefile there are some problems : 1) the -q64 option need to be added to the CFLAGS 2)the CPP needs to be the preprocessor , not the c++ compiler .. : it should be in /usr/local/IBM_compilers/xlf/9.1/etc/cpp Have fun Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 "Alex Merchant" Sent by: cpmd-list-bounces at cpmd.org 04/19/2005 08:34 AM To cc Subject [CPMD-list] Update to previous question Hi Again, I edited the Makefile pertinent to the available compilers (see http://www.vpac.org/content/systems_and_support/support/how_to/edda_start.php for details). I've made a mkefile with the following: #---------------------------------------------------------------------------- # Makefile for cpmd.x (plane wave electronic calculation) # Configuration: IBM-SP4-SMP-64 # Creation of Makefile: Apr 19 2005 # on Linux edda 2.6.5-7.145-pseries64 #1 SMP Thu Jan 27 09:19:29 UTC 2005 ppc64 ppc64 ppc64 GNU/Linux # Author: alex #---------------------------------------------------------------------------- # SHELL = /bin/sh # #--------------- Default Configuration for IBM-SP5-SMP-64 --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 CPP = /usr/local/IBM_compilers/vacpp/7.0/bin/xlc++ -P CPPFLAGS = -D__IBM -DLAPACK -DPOINTER8 -DPARALLEL=PARALLEL -DMP_LIBRARY=__MPI -DFFT_ESSL NOOPT_FLAG =CC = xlc_r FC = xlf_r -c -O3 -qstrict LD = xlf_r -O3 -qstrict AR = /usr/bin/ar #---------------------------------------------------------------------------- CFGDEST = . CFGMACH = IBM-SP5-SMP-64 # # LIST OF FILES [etc] The output I get from the compiler is: xlf_r -c -O3 -qstrict -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp ./cpmd.f -o ./cpmd.o ** cpmd_stuttgart === End of Compilation 1 === "./cpmd.f", line 25.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". "./cpmd.f", line 26.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". ** cpmd === End of Compilation 2 === 1501-511 Compilation failed for file cpmd.f. make: *** [cpmd.o] Error 1 Alex Merchant == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. _______________________________________________ CPMD-list mailing list CPMD-list at cpmd.org http://cpmd.org/mailman/listinfo/cpmd-list -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050419/a7fa3a75/attachment.html From cur at zurich.ibm.com Tue Apr 19 12:24:28 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Tue, 19 Apr 2005 12:24:28 +0200 Subject: Fw: [CPMD-list] Compilation on Power5 Message-ID: Alex, if you wait the end of the week - the new release will have makefiles and code for the power5 - Do you have AIX or PowerLinux ? With AIX is straigthforward - just take a pwr4 Makefile and substitute in -qarch pwr5 to pwr4. In the std conf you need lapack and essl Best REgards, Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 "Alex Merchant" Sent by: cpmd-list-bounces at cpmd.org 04/19/2005 08:00 AM To cc Subject [CPMD-list] Compilation on Power5 Hi All, Has anyone gotten CPMD to compile on a Power5 system? Any clues on how to edit the makefile to get it up and running? Alex Merchant == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. _______________________________________________ CPMD-list mailing list CPMD-list at cpmd.org http://cpmd.org/mailman/listinfo/cpmd-list -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050419/ff3af9d3/attachment.html From s.ketrat at uni-koeln.de Wed Apr 20 02:12:53 2005 From: s.ketrat at uni-koeln.de (Sombat Ketrat) Date: Wed, 20 Apr 2005 02:12:53 +0200 Subject: [CPMD-list] different results from P4 and Opteron Message-ID: <42659E85.7020307@uni-koeln.de> Hi CPMD users, I just recompiled the CPMD for Opteron system with pgf90 v.5.2-4 without any error. First I check the binary with the testing jobs provided by cpmd.org. For the job with use VANDERBILT PP (1,4-benzoquinone), I got the difference resulting energies and VANDERBILT AUGMENTATION CHARGES (as show below) from p4 and Opteron. From p4 VANDERBILT AUGMENTATION CHARGES (MEAN VALUE PER ATOM) ATOM TYPE NR. OF ATOMS CHARGE C 6 0.109 H 4 0.031 O 2 0.671 (K+E1+L+N+X) TOTAL ENERGY = -68.29852349 A.U. (K) KINETIC ENERGY = 36.54586290 A.U. (E1=A-S+R) ELECTROSTATIC ENERGY = -53.52384683 A.U. (S) ESELF = 57.18172686 A.U. (R) ESR = 2.52225757 A.U. (L) LOCAL PSEUDOPOTENTIAL ENERGY = -47.37768022 A.U. (N) N-L PSEUDOPOTENTIAL ENERGY = 14.54679859 A.U. (X) EXCHANGE-CORRELATION ENERGY = -18.48965791 A.U. GRADIENT CORRECTION ENERGY = -0.90231055 A.U. **************************************************************** From Opteron VANDERBILT AUGMENTATION CHARGES (MEAN VALUE PER ATOM) ATOM TYPE NR. OF ATOMS CHARGE C 6 0.113 H 4 0.030 O 2 0.654 (K+E1+L+N+X) TOTAL ENERGY = -67.89932393 A.U. (K) KINETIC ENERGY = 36.56814850 A.U. (E1=A-S+R) ELECTROSTATIC ENERGY = -53.59726876 A.U. (S) ESELF = 57.18172686 A.U. (R) ESR = 2.52225757 A.U. (L) LOCAL PSEUDOPOTENTIAL ENERGY = -47.07160624 A.U. (N) N-L PSEUDOPOTENTIAL ENERGY = 14.53283358 A.U. (X) EXCHANGE-CORRELATION ENERGY = -18.33143101 A.U. GRADIENT CORRECTION ENERGY = -0.88939362 A.U. **************************************************************** Are there anything gone wrong with my binary file. Where is the souce of the different results? I also attach the input filed output provided by cpmd.org and my calculated outputs as well. Thank you in advanced. Sombat ** -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: inp-1 Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/b12ad361/attachment.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: out-1 Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/b12ad361/attachment-0001.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: cpmd-p4-out-1 Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/b12ad361/attachment-0002.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: cpmd-opteron-out-1 Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/b12ad361/attachment-0003.cc From legend_gongzhen at hotmail.com Wed Apr 20 03:40:49 2005 From: legend_gongzhen at hotmail.com (zhen gong) Date: Wed, 20 Apr 2005 9:40:49 +0800 Subject: [CPMD-list] say hello Message-ID: <200504201543.KAA29497@iris3.simm.ac.cn> cpmd-list?? hello! With best wishes ????????????????zhen gong ????????????????legend_gongzhen at hotmail.com ????????????????????2005-04-20 From cur at zurich.ibm.com Wed Apr 20 08:56:29 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Wed, 20 Apr 2005 08:56:29 +0200 Subject: [CPMD-list] Update to previous question In-Reply-To: Message-ID: If you are running on Suse with xlf - you need to modify in prmem.F - ps -Fvsz in ps -ovsz . Moreover you have to compile control.f without optimization flags ( without -qarch and without -qsmp - they turn on optimization) - simply xlf_r -c control.f Regards, Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 "Alex Merchant" 04/20/2005 06:26 AM To cc Subject Re: [CPMD-list] Update to previous question Hi Alex, I added the -q64 bit and found cpp in /usr/local/IBM_compilers/xlf/9.1/exe/cpp not /etc The system is running SUSE not AIX. Compilation still seems to hang at control.o, but I can wait until Friday! I am having fun! A == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. >>> Alessandro Curioni 19/04/2005 >>> Alex, in your Makefile there are some problems : 1) the -q64 option need to be added to the CFLAGS 2)the CPP needs to be the preprocessor , not the c++ compiler .. : it should be in /usr/local/IBM_compilers/xlf/9.1/etc/cpp Have fun Alessandro CURIONI, PhD Research Staff Member Computational Biochemistry and Material Science group IBM Research Division - Zurich Research Laboratory Saumerstrasse 4 8003 Rueschlikon - Switzerland e-mail: cur at zurich.ibm.com www: www.zurich.ibm.com Tel: +41-1-7248633 Fax: +41-1-7248958 "Alex Merchant" Sent by: cpmd-list-bounces at cpmd.org 04/19/2005 08:34 AM To cc Subject [CPMD-list] Update to previous question Hi Again, I edited the Makefile pertinent to the available compilers (see http://www.vpac.org/content/systems_and_support/support/how_to/edda_start.php for details). I've made a mkefile with the following: #---------------------------------------------------------------------------- # Makefile for cpmd.x (plane wave electronic calculation) # Configuration: IBM-SP4-SMP-64 # Creation of Makefile: Apr 19 2005 # on Linux edda 2.6.5-7.145-pseries64 #1 SMP Thu Jan 27 09:19:29 UTC 2005 ppc64 ppc64 ppc64 GNU/Linux # Author: alex #---------------------------------------------------------------------------- # SHELL = /bin/sh # #--------------- Default Configuration for IBM-SP5-SMP-64 --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 LFLAGS = -L/usr/local/lapack -llapack -lesslsmp -llapack -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 -qsmp=omp $(QMMM_LIBS) CFLAGS = -O3 -qstrict -qarch=pwr5 CPP = /usr/local/IBM_compilers/vacpp/7.0/bin/xlc++ -P CPPFLAGS = -D__IBM -DLAPACK -DPOINTER8 -DPARALLEL=PARALLEL -DMP_LIBRARY=__MPI -DFFT_ESSL NOOPT_FLAG =CC = xlc_r FC = xlf_r -c -O3 -qstrict LD = xlf_r -O3 -qstrict AR = /usr/bin/ar #---------------------------------------------------------------------------- CFGDEST = . CFGMACH = IBM-SP5-SMP-64 # # LIST OF FILES [etc] The output I get from the compiler is: xlf_r -c -O3 -qstrict -O3 -qstrict -q64 -qmaxmem=64536 -qtune=pwr5 -qarch=pwr5 -qsmp=omp ./cpmd.f -o ./cpmd.o ** cpmd_stuttgart === End of Compilation 1 === "./cpmd.f", line 25.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". "./cpmd.f", line 26.12: 1513-039 (S) Number of arguments is not permitted for the INTRINSIC procedure "signal". ** cpmd === End of Compilation 2 === 1501-511 Compilation failed for file cpmd.f. make: *** [cpmd.o] Error 1 Alex Merchant == Dr Alex Merchant, AIP (Vic) Hon. Secretary Double Degree Coordinator Phys/Eng BP007/BP008 Lecturer, Applied Physics, RMIT phone: 9925 3388, fax: 9925 5290 mail: GPO Box 2476V, Melbourne 3001 Vic, Australia. "life ... a race between education and disaster", H.G.Wells Applied Physics is a member of the School of Applied Sciences within the SET Portfolio at RMIT. _______________________________________________ CPMD-list mailing list CPMD-list at cpmd.org http://cpmd.org/mailman/listinfo/cpmd-list -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/ec50ee0c/attachment.html From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Wed Apr 20 11:00:38 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Wed, 20 Apr 2005 11:00:38 +0200 (CEST) Subject: [CPMD-list] different results from P4 and Opteron In-Reply-To: <42659E85.7020307@uni-koeln.de> Message-ID: On Wed, 20 Apr 2005, Sombat Ketrat wrote: dear sombat, SK> Hi CPMD users, SK> SK> I just recompiled the CPMD for Opteron system with pgf90 v.5.2-4 without SK> any error. First I check the binary with the testing jobs provided by SK> cpmd.org. compiling a fortran code without any errors does not say a lot about whether the result is running correctly. but this is what test jobs are for. ;-) SK> For the job with use VANDERBILT PP (1,4-benzoquinone), I got the SK> difference resulting energies and VANDERBILT AUGMENTATION CHARGES (as SK> show below) from p4 and Opteron. in general you should not pay too much attention to that quantity as it is displayed in the cpmd output. it should be the same when running identical jobs, though. SK> Are there anything gone wrong with my binary file. Where is the souce of SK> the different results? I also attach the input filed output provided by SK> cpmd.org and my calculated outputs as well. you seem to have a miscompiled opteron binary. the pgi compilers are known for not always being fully reliable at full optimization levels. i just checked on a machine with pgi-5.2-1 installed, where i could reproduce your (wrong?) opteron numbers with optimization turned on and got your p4 numbers with -O0. using the intel compiler in 32- and 64-bit with full optimization and different libraries, i always get your p4 numbers to the last digit. regards, axel. SK> SK> Thank you in advanced. SK> Sombat SK> SK> ** SK> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From g0403127 at nus.edu.sg Wed Apr 20 11:15:59 2005 From: g0403127 at nus.edu.sg (Hong Won Keon) Date: Wed, 20 Apr 2005 17:15:59 +0800 Subject: [CPMD-list] about Makefile in cpmd2cube Message-ID: <9E4557B19235FE47B1C7B52DE3A536A33E04A1@MBOX23.stu.nus.edu.sg> Dear everybody in CPMD-list, I'm trying to make cpmd2cube executive file. But as you expected, I'm struggling with to configure Makefile. I made Makefile setting Linux-Alpha. I'm using g77 compiler. FC = g77 FFLAGS = -C -E -DFFT_DEFAULT -DLAPACK -free LFLAGS = -O LIBS = LAPACK With the above setting, when I run make, still I have an error message. How do I need to mend the configuration part in Makefile? Thanks in advance HONG won keon -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050420/666742ff/attachment.html From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Wed Apr 20 11:59:44 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Wed, 20 Apr 2005 11:59:44 +0200 (CEST) Subject: [CPMD-list] about Makefile in cpmd2cube In-Reply-To: <9E4557B19235FE47B1C7B52DE3A536A33E04A1@MBOX23.stu.nus.edu.sg> Message-ID: On Wed, 20 Apr 2005, Hong Won Keon wrote: HWK> Dear everybody in CPMD-list, dear hong, HWK> I'm trying to make cpmd2cube executive file. HWK> HWK> But as you expected, I'm struggling with to configure Makefile. HWK> HWK> I made Makefile setting Linux-Alpha. HWK> HWK> I'm using g77 compiler. this cannot work. cpmd2cube is fortran 90 code, so you need a fortran 90 compiler, i.e. the DEC/Compaq/HP/Whatever compilers. you can try your luck with g95, but i don't know if anybody ever tried it. regards, axel. HWK> HWK> HWK> HWK> FC = g77 HWK> HWK> FFLAGS = -C -E -DFFT_DEFAULT -DLAPACK -free HWK> HWK> LFLAGS = -O HWK> HWK> LIBS = LAPACK HWK> HWK> HWK> HWK> With the above setting, when I run make, still I have an error message. HWK> HWK> HWK> HWK> How do I need to mend the configuration part in Makefile? HWK> HWK> HWK> HWK> Thanks in advance HWK> HWK> HWK> HWK> HONG won keon HWK> HWK> -- ======================================================================= Dr. Axel Kohlmeyer e-mail: axel.kohlmeyer at theochem.ruhr-uni-bochum.de Lehrstuhl fuer Theoretische Chemie Phone: ++49 (0)234/32-26673 Ruhr-Universitaet Bochum - NC 03/53 Fax: ++49 (0)234/32-14045 D-44780 Bochum http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/ ======================================================================= If you make something idiot-proof, the universe creates a better idiot. From knba713 at hotmail.com Fri Apr 29 22:19:32 2005 From: knba713 at hotmail.com (Xudong Tang) Date: Fri, 29 Apr 2005 16:19:32 -0400 Subject: [CPMD-list] (no subject) Message-ID: Hello, I got a question about Geometry Optimization. I want to find out the force on each atom. Is the last column in the following output refering to the component of forces in Z direction? If so, what is the unit of these forces? "minus" means +Z direction, am I right? ATOM COORDINATES GRADIENTS (-FORCES) 1 Pt 0.0000 0.0000 0.0000 0.000E+00 0.000E+00 2.653E-02 2 Pt 5.2515 0.0000 0.0000 0.000E+00 0.000E+00 2.653E-02 3 Pt 10.5031 0.0000 0.0000 0.000E+00 0.000E+00 2.653E-02 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 25 Pt -2.6258 10.6119 8.1812 0.000E+00 0.000E+00 -1.519E-02 26 Pt 2.6258 10.6119 8.1812 0.000E+00 0.000E+00 -1.519E-02 27 Pt 7.8773 10.6119 8.1812 0.000E+00 0.000E+00 -1.519E-02 **************************************************************** *** TOTAL STEP NR. 140 GEOMETRY STEP NR. 10 *** *** GNMAX= 1.518746E-02 [3.55E-02] ETOT= -709.687796 *** *** GNORM= 7.739246E-03 DETOT= -6.802E-03 *** *** CNSTR= 0.000000E+00 TCPU= 6430.99 *** **************************************************************** Thanks. Xudong