From eesantis at unity.ncsu.edu Tue Aug 2 06:45:52 2005 From: eesantis at unity.ncsu.edu (Erik Santiso) Date: Tue, 2 Aug 2005 00:45:52 -0400 (EDT) Subject: [CPMD-list] PRFO question In-Reply-To: <20050721203556.bc6it5kv09ycc48w@my4.dal.ca> References: <20050721203556.bc6it5kv09ycc48w@my4.dal.ca> Message-ID: <2404.200.90.11.144.1122957952.squirrel@webmail.ncsu.edu> Hi all, I have (another) question about starting a PRFO calculation using a Hessian from a previous geometry optimization. In this case I have some atoms in the system that are fixed (with FIX ATOMS). If I run a vibrational analysis for such a system, CPMD (version 3.9.1) does not generate a HESSIAN file. In order to create one, I re-run the vibrational analysis with the constraints removed and RESTART VIBANALYSIS. The problem is, if I then use this Hessian to start a P-RFO calculation (with RESTART HESSIAN and PRFO HESSTYPE 1, and putting the constraints back in), CPMD complains: HESSIN | NUMBER OF PARAMETERS NOT CONSISTENT 234 42 (I have 14 movable atoms and 64 fixed ones) HESSIN | BAD HESSIAN FILE ./HESSIAN and goes on trying to generate an empirical Hessian etc. I've tried using RESTART PHESS and HESSIAN PARTIAL, but CPMD complains that there is no partial Hessian information on the restart file. So the question is: is there a way to either (1) Get CPMD to generate a Hessian file with the correct info from a vibrational analysis when some of the atoms are fixed (I thought of just changing the number of parameters in the header of the HESSIAN file, but a little voice in my head says this may not be a good idea) or (2) Get CPMD to copy the right part of the Hessian onto the PRFO partial Hessian? The truth is that I could just let CPMD calculate the Hessian again (with PRFO HESSTYPE 0), but it seems a waste of computing time... Sorry for the long question, and many thanks for any suggestions! Erik. --------------------------------------- A train station is where a train stops. A bus station is where a bus stops. My computer is a workstation... From wkhan at wow.hongik.ac.kr Tue Aug 2 23:00:04 2005 From: wkhan at wow.hongik.ac.kr (Wone Keun Han) Date: Tue, 2 Aug 2005 17:00:04 -0400 Subject: [CPMD-list] Help, compiling problem! Message-ID: <000e01c597a5$2aa87bd0$9d3c9480@physics.brown.edu> Dear Shyma and Members, The Linux seems to recognize the single cpu PC as two processor PC if it has dual channel memory. I mean LINUX-SMP can be installed. I am wondering if we can speed up CPMD by compiling it as SMP compared to serial calculation. If so, I would like to have more information. Thanks, Wone Keun Han -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050802/5d5fb0d3/attachment.html From krekeler at mpip-mainz.mpg.de Wed Aug 3 13:57:45 2005 From: krekeler at mpip-mainz.mpg.de (Christian Krekeler) Date: Wed, 03 Aug 2005 13:57:45 +0200 Subject: [CPMD-list] triclinic symmetry for disordered system Message-ID: <42F0B139.6000109@mpip-mainz.mpg.de> Hi, I got this output when I tried to dress up one test configuration of 100 water molecules (SPCE modell) in a triclinic box with density for some further calculations. This should be a test for NaCl in water with the ions along the diagonal. Could it be that this problem occurs due to the use of a disordered system in combination with the triclinic symmetry? Can it possible be fixed? Thanks in advance! regards Christian Krekeler *** LOADPA| THE NEW SIZE OF THE PROGRAM IS 10384 KBytes *** PROC= 1 IG= 13 HG(IG)= 17.5003600288021559 HG(IG-1)= 17.5003600864090672 PROC= 1 IG= 277 HG(IG)= 124.001440093606021 HG(IG-1)= 124.001440345636240 PROC= 1 IG= 442 HG(IG)= 172.010080864067334 HG(IG-1)= 172.010080957678582 PROC= 1 IG= 535 HG(IG)= 197.507560806469485 HG(IG-1)= 197.507561036897130 PROC= 1 IG= 821 HG(IG)= 268.004320288018960 HG(IG-1)= 268.004320842485470 PROC= 1 IG= 855 HG(IG)= 276.014401900976225 HG(IG-1)= 276.014402484246204 PROC= 1 IG= 996 HG(IG)= 305.999999999999943 HG(IG-1)= 306.000001036924289 PROC= 1 IG= 998 HG(IG)= 305.999999999999943 HG(IG-1)= 306.000001036924345 PROC= 2 IG= 53 HG(IG)= 41.5006000432030149 HG(IG-1)= 41.5006002160237557 PROC= 2 IG= 79 HG(IG)= 57.5021601584112219 HG(IG-1)= 57.5021603888388668 PROC= 2 IG= 98 HG(IG)= 67.9999999999999716 HG(IG-1)= 68.0000001800215870 PROC= 2 IG= 217 HG(IG)= 108.008400684051807 HG(IG-1)= 108.008400856872555 PROC= 2 IG= 400 HG(IG)= 156.003360230415495 HG(IG-1)= 156.003360554454417 PROC= 2 IG= 451 HG(IG)= 172.001680108006866 HG(IG-1)= 172.001680453648333 PROC= 2 IG= 469 HG(IG)= 177.999999999999943 HG(IG-1)= 178.000000180021544 PROC= 2 IG= 518 HG(IG)= 195.999999999999943 HG(IG-1)= 196.000000352842278 PROC= 2 IG= 522 HG(IG)= 196.009600756056130 HG(IG-1)= 196.009601036889791 PROC= 2 IG= 706 HG(IG)= 237.496759870396033 HG(IG-1)= 237.496760388858263 PROC= 2 IG= 798 HG(IG)= 260.000000064807693 HG(IG-1)= 260.000000352842278 PROC= 2 IG= 800 HG(IG)= 260.000000583269923 HG(IG-1)= 260.000000871304508 PROC= 2 IG= 840 HG(IG)= 268.002400172812031 HG(IG-1)= 268.002400324030191 PROC= 3 IG= 65 HG(IG)= 53.5018001728142423 HG(IG-1)= 53.5018002880280648 PROC= 3 IG= 233 HG(IG)= 121.499399971198656 HG(IG-1)= 121.499400144019404 PROC= 3 IG= 549 HG(IG)= 208.011520950472544 HG(IG-1)= 208.011521152096719 PROC= 3 IG= 850 HG(IG)= 275.997599884794738 HG(IG-1)= 275.997600576077730 PROC= 3 IG= 1157 HG(IG)= 340.000000352842221 HG(IG-1)= 340.000000871304451 PROC= 3 IG= 1410 HG(IG)= 388.001920122407682 HG(IG-1)= 388.001920576062105 PROC= 3 IG= 2110 HG(IG)= 507.998559920795458 HG(IG-1)= 507.998559978402398 PROC= 3 IG= 2121 HG(IG)= 508.046804025114454 HG(IG-1)= 508.046804428362748 PROC= 3 IG= 2308 HG(IG)= 538.020161562514772 HG(IG-1)= 538.020162621041777 PROC= 3 IG= 2582 HG(IG)= 580.000001216945748 HG(IG-1)= 580.000002081049502 PROC= 3 IG= 2666 HG(IG)= 592.003840259217100 HG(IG-1)= 592.003840691268920 PROC= 3 IG= 3242 HG(IG)= 676.025201864932910 HG(IG-1)= 676.025203132285014 PROC= 3 IG= 3593 HG(IG)= 723.989439481575573 HG(IG-1)= 723.989440237666258 PROC= 4 IG= 28 HG(IG)= 28.0014401152086378 HG(IG-1)= 28.0014401512129574 PROC= 4 IG= 67 HG(IG)= 52.0014401152086307 HG(IG-1)= 52.0014403456362757 [......] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 0] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 1] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 3] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 4] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 5] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 6] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 7] PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 2] -- ---------------------------------------------------------- Christian Krekeler Max Planck Institute for Polymer Research Ackermannweg 10 D-55128 Mainz, Germany Phone: +49 6131 379 148 Fax: +49 6131 379 340 E-mail: krekeler at mpip-mainz.mpg.de web: www.mpip-mainz.mpg.de/theory.html ---------------------------------------------------------- From carla.molteni at kcl.ac.uk Fri Aug 5 19:00:37 2005 From: carla.molteni at kcl.ac.uk (Dr Carla Molteni) Date: Fri, 5 Aug 2005 18:00:37 +0100 Subject: [CPMD-list] Ph D studentship at King's College London Message-ID: <1123261237.42f39b35f215f@impmail.kcl.ac.uk> Applications are invited for a Ph.D. studentship, available for UK citizens, to work on ab initio simulations of nanomaterials at King's College London, (Physics Department). Further details on projects and application procedure can be obtained from Dr Carla Molteni (carla.molteni at kcl.ac.uk) and from the Graduate Tutor Dr David Richards (david.r.richards at kcl.ac.uk). Best regards, Carla Molteni -- Dr Carla Molteni Tel. +44 20 7848 2170 Physics Department Fax. +44 20 7848 2420 King's College London email carla.molteni at kcl.ac.uk Strand London WC2R 2LS (UK) From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Tue Aug 9 12:45:13 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Tue, 9 Aug 2005 12:45:13 +0200 (CEST) Subject: [CPMD-list] Help, compiling problem! In-Reply-To: <000e01c597a5$2aa87bd0$9d3c9480@physics.brown.edu> Message-ID: On Tue, 2 Aug 2005, Wone Keun Han wrote: WKH> Dear Shyma and Members, WKH> WKH> WKH> WKH> The Linux seems to recognize the single cpu PC as two processor PC if WKH> it has dual channel memory. WKH> WKH> I mean LINUX-SMP can be installed. you _still_ have only one processor. what is the point? you would have two threads competing for the same cpu, the same memory and the same floating-point unit(!). on top of that, you have the overhead of compiling in the thread management for openmp. WKH> WKH> I am wondering if we can speed up CPMD by compiling it as SMP compared to WKH> serial calculation. based on the facts listed above, CPMD will become (much) slower. in fact, i made quite a few tests on real multiprocessor PC hardware and even there i found little evidence that suggested the use of openmp. due to the external memory controller and the extremely high clock rate, intel xeon cpu are worst, dual athlon (32-bit) are a bit better, but in both cases the overhead for OpenMP is so high, that you are either better off using two MPI threads per SMP node or not using the second cpu AT ALL. thanks to the internal memory controller opteron cpus are a different story, but again two MPI threads is much more efficient than using OpenMP (unless you run across are really, really largen number of nodes and even then i found a higher speed gain using other tricks, e.g. TASKGROUPS than with OpenMP). the best openmp performance on SMP machines, so far, i found on with the power4/5 architecture (either on ibm hardware or on mac osx machines using the ibm compiler). also using OpenMP is essential for many vector machines, but that is a different story. WKH> WKH> If so, I would like to have more information. hope this will satisfy you. you can find some (pretty old by now) benchmark numbers and discussions on http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/cpmd-bench.html regards, axel. WKH> WKH> WKH> WKH> Thanks, WKH> WKH> WKH> WKH> Wone Keun Han WKH> WKH> WKH> WKH> -- ======================================================================= 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 lyuan at Princeton.EDU Fri Aug 12 05:51:44 2005 From: lyuan at Princeton.EDU (Lan-Feng Yuan (lyuan@Princeton.EDU)) Date: Thu, 11 Aug 2005 23:51:44 -0400 Subject: [CPMD-list] wrong velocities in TRAJECTORY? Message-ID: Hi, dear CPMD users, I want to calculate the kinetic energies of each atom to check equipartitioning. The atomic velocites are stored in the TRAJECTORY file. But in doing that, I am surprised to see that many of the velocites are wrong in path integral calculations! For example, it has lines like: 1 6.82425508971048 9.52450269497126 9.51015987345505 0.00000000000000 0.00000000000000 0.00000 000000000 1 12.21231276874373 9.33303372434401 9.34606460682084 0.00000000000000 0.00000000000000 0.00000 000000000 1 6.00617010855165 11.38389016650084 8.59588245596876 0.00000000000000 0.00000000000000 0.00000 000000000 1 6.13293595832137 8.15809729569046 8.91367284452267 0.00000000000000 0.00000000000000 0.00000 000000000 1 10.10949129407331 9.84789884632727 9.34621236955664 0.00000000000000 0.00000000000000 0.00000 000000000 1 13.21957008182102 9.26567361368209 11.23069151637693 0.00000000000000 0.00000000000000 0.00000 000000000 ... 6000 5.53260233231064 10.68899396835529 8.78025724226381 0.00000000000000 0.00000000000000 0.00000 000000000 6000 5.59725546928881 8.65292224670147 10.21791288576526 0.00000000000000 0.00000000000000 0.00000 000000000 6000 10.33714449967338 10.39804084083272 9.61533970021246 0.00000000000000 0.00000000000000 0.00000 000000000 6000 12.35422745525717 9.05625092206057 11.31325603410190 0.00000000000000 0.00000000000000 0.00000 000000000 Namely, some beads are reported to have zero velocites from the first step to the last step, but this is obviously not so, because we can see that the atomic positions have changed. And when I restart a new calculation from the RESTART files, it began to report non-zero velocites, but still those velocities don't change with steps. I do hope this is a small bug only affecting this specific output ... On the other hand, is there any other way to get velocities? Thank you very much! Best regards, Lan-Feng ---------------------------- Dr. Lan-Feng Yuan, Postdoctoral Associate, Department of Chemistry, Princeton University, Princeton, NJ 08544, USA Office Phone: 1-609-258-1834 Fax: 1-609-258-6746 ---------------------------- From jasanc0 at engr.uky.edu Tue Aug 9 18:53:49 2005 From: jasanc0 at engr.uky.edu (Jaime Sanchez) Date: Tue, 9 Aug 2005 12:53:49 -0400 Subject: [CPMD-list] compile cpmd2cube Message-ID: <200508120655.IAA08390@internet-fence.zurich.ihost.com> Dear CPMD users, I want to compile cpmd2cube for the configuration HPUX-IA64 but this configuration is not available when I type ./Configure. Anyone can help me with this? Thanks Jaime -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050809/7b873afb/attachment.html From gn203 at cam.ac.uk Fri Aug 12 08:59:14 2005 From: gn203 at cam.ac.uk (Gana Natarajan) Date: Fri, 12 Aug 2005 07:59:14 +0100 Subject: [CPMD-list] Constant pressure with BO dynamics Message-ID: <42FC48C2.5000309@cam.ac.uk> Dear list members, I would like to perform a constant-pressure molecular dynamics run using Born-Oppenheimer dynamics, could somebody tell me if it is possible. Thank you. Regards, Gana From gn203 at cam.ac.uk Fri Aug 12 09:11:33 2005 From: gn203 at cam.ac.uk (Gana Natarajan) Date: Fri, 12 Aug 2005 08:11:33 +0100 Subject: [CPMD-list] Constant pressure MD with BO dynamics Message-ID: <42FC4BA5.7000707@cam.ac.uk> Dear List Members, I would like to know how to perform constant-pressure molecular dynamics with Born-Oppenheimer dynamics. Thanks, Gana From shyma at imr.ac.cn Fri Aug 12 09:19:08 2005 From: shyma at imr.ac.cn (=?gb2312?B?wu3J0NLl?=) Date: Fri, 12 Aug 2005 15:19:08 +0800 (CST) Subject: [CPMD-list] Help:about ag cutoff! Message-ID: <42FC4D6C.000178.12880@app-01> Dear CPMD user: I want to specify the suitable Ag cutoff energy and compute the single point total energy with different cutoff, I find it's difficult to converge with the criteria 1d-6 no matter what arithmetic I use. The Ag pesudopential file is Ag_SEMI_BLYP.psp under the pseudo-ext folder. The table show the results: Ag_SEMI_BLYP.psp a=4.086 Å converge criteria 1d-6 * indate it does't converge when it reach the maxmum number step ??????????????A.U.?? (the number of computation step time/s) CUTOFF PCG MINIMIZE DEFAULT LANCZOS DIAGONALIZATION STEEPEST DESCEND ODIIS- LANCZOS 20 -148.65459706 (2521 129s) -148.62911313* ( 234s) -148.62911313* ( 234s) -148.63166745* -148.65454618* ( 238s) 30 -150.48724701 (902 73s) -150.48724701 (2529 103s) -150.48724701 (2529 103s) -150.48724593* -150.48694426* ( 411s) 40 -150.58035561 (2986 306) -150.58035560 (5742 281s) -150.58035560 (5742 281s) -150.58009052* -150.58008550* ( 495s) 50 -150.58300390 (2166 363) -150.58272049 (5065 399s) -150.58272049 (5065 399s) -150.58272548* -150.58300309* ???? 60 -150.58499425 (2231 408) -150.58471114 ??5679 490s?? -150.58471114 (5679 493s) -150.58493776* -150.58480024* ( 880s) 70 -150.58628042 (1958 449) -150.58622285* ( 1126s) -150.58622285* ( 1126s) -150.58627646* 80 -150.58682224 (2060 627s) -150.58682223 (1649 244s) -150.58682223 (1649 245s) -150.58681380* 90 -150.58693002 (6863 2500s) -150.58665354* ( 1605s) -150.58665354* ( 1615s) -150.58665277* 100 -150.58697361 (4912 2619s) -150.58671318* ( 1939s) -150.58680340* ( 1934) 110 -150.58699188 (1647 707s) -150.58699187 (5925 1248s) -150.58699187 (5925 1253s) -150.58672231 ( 2092s) 120 -150.58701175 (2112 1342s) -150.58672938* ( 3316s) -150.58680001* 130 -150.58701943 (644 433s) -150.58701942 (77 27s) -150.58701942 (77 26s) -150.58701942 (6653 ) 140 -150.58702605 (3426 2472s) -150.58702605 (4539 1588s) -150.58700999* 150 -150.58703046 (1468 1187s) -150.58674744 (5322 2050s) -150.58684764* ( 3830s) And you can find that the computation time is very short if I specify the cutoff 130 Ry, but it's difficult to converge in other casese! The following is my main form of input file: &CPMD OPTIMIZE WAVEFUNCTION CONVERGENCE ORBITALS 1.0D-6 PRINT FORCES ON PCG MINIMIZE TIMESTEP 20 MIRROR &END &DFT FUNCTIONAL BLYP &END &SYSTEM SYMMETRY FCC SCALE ANGSTROM CELL 4.086 1.0 1.0 0 0 0 CUTOFF 20 &END &ATOMS *Ag_SEMI_BLYP.psp LMAX=D 1 0.00 0.00 0.00 &END I appreciate that if you find something wrong with the input file and it should be corrected or if you have some advice to improve the effiency of the computation! Thank in advance and best wishes! shyma -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050812/d7912b37/attachment.html From alex.merchant at rmit.edu.au Fri Aug 12 09:36:46 2005 From: alex.merchant at rmit.edu.au (Alex Merchant) Date: Fri, 12 Aug 2005 17:36:46 +1000 Subject: [CPMD-list] Re: CPMD-list Digest, Vol 16, Issue 2 (RE: Your Message) Message-ID: Hi, I am on leave until Monday 22nd 2005. During this period I will not be checking my email or phone messages. If you have urgent course issues, please contact Applied Physics on (03) 9925 2600 and speak to the Service Physics Coordinator Peter Daivis. Eratosthenes Project - Australia please go to the website www.rmit.edu.au/scienceweek for support contact details during Science Week. For issues about Double Degrees admissions please contact Ken McGregor in Physics, for Academic Issues please contact Ian Palmer in Engineering or Salvy Russo in Physics. Yours truly Alex Merchant >>> cpmd-list 08/12/05 17:21 >>> Send CPMD-list mailing list submissions to cpmd-list at cpmd.org To subscribe or unsubscribe via the World Wide Web, visit http://cpmd.org/mailman/listinfo/cpmd-list or, via email, send a message with subject or body 'help' to cpmd-list-request at cpmd.org You can reach the person managing the list at cpmd-list-owner at cpmd.org When replying, please edit your Subject line so it is more specific than "Re: Contents of CPMD-list digest..." From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Tue Aug 9 12:48:19 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Tue, 9 Aug 2005 12:48:19 +0200 (CEST) Subject: [CPMD-list] triclinic symmetry for disordered system In-Reply-To: <42F0B139.6000109@mpip-mainz.mpg.de> Message-ID: On Wed, 3 Aug 2005, Christian Krekeler wrote: CK> Hi, CK> I got this output when I tried to dress up one test configuration of 100 CK> water CK> molecules (SPCE modell) in a triclinic box with density for some further CK> calculations. CK> This should be a test for NaCl in water with the ions along the diagonal. hi, could you please post your input (w/o the coordinates). i don't quite get what you are aiming at. regards, axel. CK> CK> Could it be that this problem occurs due to the use of a disordered CK> system in combination with the triclinic symmetry? Can it possible be fixed? CK> CK> Thanks in advance! CK> CK> regards CK> Christian Krekeler CK> CK> *** LOADPA| THE NEW SIZE OF THE PROGRAM IS 10384 KBytes *** CK> PROC= 1 IG= 13 HG(IG)= 17.5003600288021559 HG(IG-1)= 17.5003600864090672 CK> PROC= 1 IG= 277 HG(IG)= 124.001440093606021 HG(IG-1)= CK> 124.001440345636240 CK> PROC= 1 IG= 442 HG(IG)= 172.010080864067334 HG(IG-1)= CK> 172.010080957678582 CK> PROC= 1 IG= 535 HG(IG)= 197.507560806469485 HG(IG-1)= CK> 197.507561036897130 CK> PROC= 1 IG= 821 HG(IG)= 268.004320288018960 HG(IG-1)= CK> 268.004320842485470 CK> PROC= 1 IG= 855 HG(IG)= 276.014401900976225 HG(IG-1)= CK> 276.014402484246204 CK> PROC= 1 IG= 996 HG(IG)= 305.999999999999943 HG(IG-1)= CK> 306.000001036924289 CK> PROC= 1 IG= 998 HG(IG)= 305.999999999999943 HG(IG-1)= CK> 306.000001036924345 CK> PROC= 2 IG= 53 HG(IG)= 41.5006000432030149 HG(IG-1)= 41.5006002160237557 CK> PROC= 2 IG= 79 HG(IG)= 57.5021601584112219 HG(IG-1)= 57.5021603888388668 CK> PROC= 2 IG= 98 HG(IG)= 67.9999999999999716 HG(IG-1)= 68.0000001800215870 CK> PROC= 2 IG= 217 HG(IG)= 108.008400684051807 HG(IG-1)= CK> 108.008400856872555 CK> PROC= 2 IG= 400 HG(IG)= 156.003360230415495 HG(IG-1)= CK> 156.003360554454417 CK> PROC= 2 IG= 451 HG(IG)= 172.001680108006866 HG(IG-1)= CK> 172.001680453648333 CK> PROC= 2 IG= 469 HG(IG)= 177.999999999999943 HG(IG-1)= CK> 178.000000180021544 CK> PROC= 2 IG= 518 HG(IG)= 195.999999999999943 HG(IG-1)= CK> 196.000000352842278 CK> PROC= 2 IG= 522 HG(IG)= 196.009600756056130 HG(IG-1)= CK> 196.009601036889791 CK> PROC= 2 IG= 706 HG(IG)= 237.496759870396033 HG(IG-1)= CK> 237.496760388858263 CK> PROC= 2 IG= 798 HG(IG)= 260.000000064807693 HG(IG-1)= CK> 260.000000352842278 CK> PROC= 2 IG= 800 HG(IG)= 260.000000583269923 HG(IG-1)= CK> 260.000000871304508 CK> PROC= 2 IG= 840 HG(IG)= 268.002400172812031 HG(IG-1)= CK> 268.002400324030191 CK> PROC= 3 IG= 65 HG(IG)= 53.5018001728142423 HG(IG-1)= 53.5018002880280648 CK> PROC= 3 IG= 233 HG(IG)= 121.499399971198656 HG(IG-1)= CK> 121.499400144019404 CK> PROC= 3 IG= 549 HG(IG)= 208.011520950472544 HG(IG-1)= CK> 208.011521152096719 CK> PROC= 3 IG= 850 HG(IG)= 275.997599884794738 HG(IG-1)= CK> 275.997600576077730 CK> PROC= 3 IG= 1157 HG(IG)= 340.000000352842221 HG(IG-1)= CK> 340.000000871304451 CK> PROC= 3 IG= 1410 HG(IG)= 388.001920122407682 HG(IG-1)= CK> 388.001920576062105 CK> PROC= 3 IG= 2110 HG(IG)= 507.998559920795458 HG(IG-1)= CK> 507.998559978402398 CK> PROC= 3 IG= 2121 HG(IG)= 508.046804025114454 HG(IG-1)= CK> 508.046804428362748 CK> PROC= 3 IG= 2308 HG(IG)= 538.020161562514772 HG(IG-1)= CK> 538.020162621041777 CK> PROC= 3 IG= 2582 HG(IG)= 580.000001216945748 HG(IG-1)= CK> 580.000002081049502 CK> PROC= 3 IG= 2666 HG(IG)= 592.003840259217100 HG(IG-1)= CK> 592.003840691268920 CK> PROC= 3 IG= 3242 HG(IG)= 676.025201864932910 HG(IG-1)= CK> 676.025203132285014 CK> PROC= 3 IG= 3593 HG(IG)= 723.989439481575573 HG(IG-1)= CK> 723.989440237666258 CK> PROC= 4 IG= 28 HG(IG)= 28.0014401152086378 HG(IG-1)= 28.0014401512129574 CK> PROC= 4 IG= 67 HG(IG)= 52.0014401152086307 HG(IG-1)= 52.0014403456362757 CK> [......] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 0] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 1] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 3] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 4] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 5] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 6] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 7] CK> CK> CK> PROGRAM STOPS IN SUBROUTINE RGGEN| BAD REORDERING [PROC= 2] CK> CK> CK> -- ======================================================================= 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 yangzy at lnm.imech.ac.cn Wed Aug 10 17:06:54 2005 From: yangzy at lnm.imech.ac.cn (Yang Zhenyu) Date: Wed, 10 Aug 2005 23:06:54 +0800 Subject: [CPMD-list] [help!]Problem on running test! Message-ID: <200508120833.KAA09054@internet-fence.zurich.ihost.com> Dear CPMD users: I am a newbie for CPMD. I have just installed CPMD on my Linux computer.And I tried to run a test, getting a problem as follows: *** DETSP| THE NEW SIZE OF THE PROGRAM IS 1628/ 49320 kBYTES *** RECPNEW| ECPFILE NOT FOUND SI_MT_LDA PROGRAM STOPS IN SUBROUTINE RECPNEW| Where can I find the needed file? It seems that some more packages need to be installed?? Anybody could enlighten me on this? ?????????????? ??Best Regards ??Yang Zhenyu State Key Laboratory of Nonlinear Mechanics (LNM) Institute of Mechanics, Chinese Academy of Sciences No.15 Beisihuanxi Road Beijing 100080, P. R. China ??yangzy at lnm.imech.ac.cn ??2005-08-10 From yangzy at lnm.imech.ac.cn Thu Aug 11 16:20:05 2005 From: yangzy at lnm.imech.ac.cn (Yang Zhenyu) Date: Thu, 11 Aug 2005 22:20:05 +0800 Subject: [CPMD-list] Can not find library files!!! Message-ID: <200508120834.KAA09218@internet-fence.zurich.ihost.com> Dear CPMD users: I am a newbie on CPMD. I have just installed CPMD on my Linux computer.And I tried to run a test, getting a problem as follows: ------------------------------------------------------------------------- *** DETSP| THE NEW SIZE OF THE PROGRAM IS 1628/ 49320 kBYTES *** RECPNEW| ECPFILE NOT FOUND SI_MT_LDA PROGRAM STOPS IN SUBROUTINE RECPNEW| ------------------------------------------------------------------------- I have tried two methods - copy the file SI_MT_LDA into your working directory _or_ - set PP_LIBRARY_PATH to the _full_ path of your pseudopotential directory (by export PP_LIBRARY_PATH=/home/yzy/CPMD-3.9.2/PPLIBNEW ) And when I run the test again, nothing changed. I am so confued. Enybody could enlighten me on this problem?? Thanks in advance. ?????????????? ??Best Regards ??Yang Zhenyu State Key Laboratory of Nonlinear Mechanics (LNM) Institute of Mechanics, Chinese Academy of Sciences No.15 Beisihuanxi Road Beijing 100080, P. R. China ??yangzy at lnm.imech.ac.cn ??2005-08-11 From bala at jncasr.ac.in Fri Aug 12 13:42:34 2005 From: bala at jncasr.ac.in (Balasubramanian Sundaram) Date: Fri, 12 Aug 2005 17:12:34 +0530 Subject: [CPMD-list] Coredump on SP4 Message-ID: <42FC8B2A.30608@jncasr.ac.in> Hello List Members, We are facing a difficulty in running cpmd on an IBM SP4. Here are the details: Version: CPMD-3.9.2 (May 14th version on cpmd.org site) The code compiles fine with the following header of the Makefile. #--------------- Default Configuration for IBM-SP4-SMP-64 --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -q64 -qtune=pwr4 -qarch=pwr4 -qsmp=omp LFLAGS = -L/usr/local/lib -llapack64 -lesslsmp -llapack64 -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr4 -qsmp=omp $(QMMM_LIBS) CFLAGS = $(FFLAGS) CPP = /usr/ccs/lib/cpp -P CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL=PARALLEL -DFFT_ESSL -DMP_LIBRARY=__MPI NOOPT_FLAG = CC = xlc_r -c FC = mpxlf_r -c LD = mpxlf_r AR = /usr/bin/ar #---------------------------------------------------------------------------- However, when we run the Si64-10ryd example (inp.wf) downloaded from Axel's homepage, it coredumps. The same example runs fine on an executable compiled with 3.9.1 version, with a similar Makefile. Pls. find attached the output of the cpmd-3.9.2 run. We appreciate any help in this regard, Many Thanks! Bala -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: file.out Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050812/f3c7762c/attachment.cc From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Fri Aug 12 14:14:19 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Fri, 12 Aug 2005 14:14:19 +0200 (CEST) Subject: [CPMD-list] Constant pressure MD with BO dynamics In-Reply-To: <42FC4BA5.7000707@cam.ac.uk> Message-ID: On Fri, 12 Aug 2005, Gana Natarajan wrote: GN> Dear List Members, dear gana, GN> I would like to know how to perform constant-pressure molecular GN> dynamics with Born-Oppenheimer dynamics. if you want to do this with CPMD, you'd have to implement it. currently, there constant pressure MD support is only implemented for car-parrinello MD. regards, axel. GN> GN> Thanks, GN> GN> Gana GN> _______________________________________________ GN> CPMD-list mailing list GN> CPMD-list at cpmd.org GN> http://cpmd.org/mailman/listinfo/cpmd-list GN> GN> -- ======================================================================= 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 Fri Aug 12 14:16:25 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Fri, 12 Aug 2005 14:16:25 +0200 (CEST) Subject: [CPMD-list] [help!]Problem on running test! In-Reply-To: <200508120833.KAA09054@internet-fence.zurich.ihost.com> Message-ID: On Wed, 10 Aug 2005, Yang Zhenyu wrote: > Dear CPMD users: > > I am a newbie for CPMD. > I have just installed CPMD on my Linux computer.And I tried to run > a test, getting a problem as follows: > > *** DETSP| THE NEW SIZE OF THE PROGRAM IS 1628/ 49320 kBYTES *** > RECPNEW| ECPFILE NOT FOUND > SI_MT_LDA hi, what version of CPMD are you using? this looks like an error we fixed quite some time ago. regards, axel. > > PROGRAM STOPS IN SUBROUTINE RECPNEW| > > > Where can I find the needed file? > It seems that some more packages need to be installed?? > Anybody could enlighten me on this? > > ?????????????? > ??Best Regards > > ??Yang Zhenyu > State Key Laboratory of Nonlinear Mechanics (LNM) > Institute of Mechanics, Chinese Academy of Sciences > No.15 Beisihuanxi Road > Beijing 100080, P. R. China > ??yangzy at lnm.imech.ac.cn > ??2005-08-10 > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list > > -- ======================================================================= 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 Fri Aug 12 14:29:10 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Fri, 12 Aug 2005 14:29:10 +0200 (CEST) Subject: [CPMD-list] Coredump on SP4 In-Reply-To: <42FC8B2A.30608@jncasr.ac.in> Message-ID: On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: BS> Hello List Members, BS> dear bala, BS> We are facing a difficulty in running cpmd on an IBM SP4. Here are BS> the details: you could try the corresponding PWR5 config and then just replace pwr5 with pwr4, but also you should probably compile the files control.o memory.o and freem.o without optimization and openmp support (-O0 -qsmp-noopt). regards, axel. BS> homepage, it coredumps. The same example runs fine on an executable BS> compiled with 3.9.1 version, with a similar Makefile. was that executable compiled on the same machine? with the same compilers? recently? also it might be helpful, if you could create a stack trace of the core dump, so we now in what subroutine it dies. BS> Pls. find attached the output of the cpmd-3.9.2 run. BS> BS> We appreciate any help in this regard, BS> BS> Many Thanks! BS> Bala BS> -- ======================================================================= 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 eneritz.muguruza at kcl.ac.uk Fri Aug 12 16:48:10 2005 From: eneritz.muguruza at kcl.ac.uk (Eneritz Muguruza) Date: Fri, 12 Aug 2005 15:48:10 +0100 Subject: [CPMD-list] ROKS and TDDFT of spin polarized systems Message-ID: <1123858090.42fcb6aa4ebd9@impmail.kcl.ac.uk> Hello everyone, Here I ask another question. I am studying the isomerization of a chromophore of a photoreceptor protein. I am doing that by calculating potential energy surfaces of the ground and first excited singlet state along the isomerization coordinate, the torsional angle around a double bond. I am calculating the excited states with ROKS and TDDFT. At the twisted angle of 90 degrees in the ground state the double bond breaks and a spin unrestricted (LSD) calculation is necessary. My question is: how can I calculate the excited state of that structure? - I tried to do the calculation with TDDFT but only a mixed state was obtained. Is there the spin-unrestricted TDDFT implemented on CPMD? - ROKS is a spin restricted method, but could I somehow do different calculatiosn by transfering the electron at HOMO with spin-up and spin-down and then do a linear combination to get the right spin? Thank you very much for any hints or comments. Regards, Eneritz -- Eneritz Muguruza eneritz.muguruza at kcl.ac.uk From ari.p.seitsonen at iki.fi Fri Aug 12 18:01:08 2005 From: ari.p.seitsonen at iki.fi (Ari P Seitsonen) Date: Fri, 12 Aug 2005 18:01:08 +0200 (CEST) Subject: [CPMD-list] Coredump on SP4 In-Reply-To: <42FC8B2A.30608@jncasr.ac.in> References: <42FC8B2A.30608@jncasr.ac.in> Message-ID: Dear Bala, In addition to Axel's suggestions, have you removed the option '-DPOINTER8' from the 'Makefile'? Greetings from Lausanne, apsi -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ CNRS & IMPMC, Universit? Pierre et Marie Curie 4 place Jussieu, case 115 / F-75252 Paris Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: > Hello List Members, > > We are facing a difficulty in running cpmd on an IBM SP4. Here are the > details: > > Version: CPMD-3.9.2 (May 14th version on cpmd.org site) > > The code compiles fine with the following header of the Makefile. > #--------------- Default Configuration for IBM-SP4-SMP-64 --------------- > SRC = . > DEST = . > BIN = . > #QMMM_FLAGS = -D__QMECHCOUPL > #QMMM_LIBS = -L. -lmm > FFLAGS = -q64 -qtune=pwr4 -qarch=pwr4 -qsmp=omp > LFLAGS = -L/usr/local/lib -llapack64 -lesslsmp -llapack64 -q64 \ > -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr4 > -qsmp=omp $(QMMM_LIBS) > CFLAGS = $(FFLAGS) > CPP = /usr/ccs/lib/cpp -P > CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL=PARALLEL -DFFT_ESSL > -DMP_LIBRARY=__MPI > NOOPT_FLAG = > CC = xlc_r -c > FC = mpxlf_r -c > LD = mpxlf_r > AR = /usr/bin/ar > #---------------------------------------------------------------------------- > > However, when we run the Si64-10ryd example (inp.wf) downloaded from Axel's > homepage, it coredumps. The same example runs fine on an executable compiled > with 3.9.1 version, with a similar Makefile. > > Pls. find attached the output of the cpmd-3.9.2 run. > > We appreciate any help in this regard, > > Many Thanks! > Bala > From bala at jncasr.ac.in Sat Aug 13 08:46:58 2005 From: bala at jncasr.ac.in (Balasubramanian Sundaram) Date: Sat, 13 Aug 2005 12:16:58 +0530 Subject: [CPMD-list] Coredump on SP4 In-Reply-To: References: Message-ID: <42FD9762.6030901@jncasr.ac.in> Dear Alessandro, Thanks for your reply. Alessandro: The following is the input file. ============ &CPMD OPTIMIZE WAVEFUNCTION MAXSTEP 50 ODIIS 15 TIMESTEP 15.0 MAXCPUTIME 3600 SPLINE POINTS 1000 &END &SYSTEM SYMMETRY 1 CELL 20.5224280754 1.0 1.0 0.0 0.0 0.0 CUTOFF 10 DUAL 3 TESR 2 2 2 SCALED &END &ATOMS *SI_SGS KLEINMAN-BYLANDER RAGGIO=1.0 LMAX=P LOC=P 63 ------>>>Coordinates skipped &DFT FUNCTIONAL LDA &END ============================= It appears that the CELL information is consistent with what you have written (20.52 1.0 1.0 0.0 0. 0.). Regards, Bala Axel Kohlmeyer wrote: > On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: > > BS> Hello List Members, > BS> > > > dear bala, > > BS> We are facing a difficulty in running cpmd on an IBM SP4. Here are > BS> the details: > > you could try the corresponding PWR5 config and then just > replace pwr5 with pwr4, but also you should probably > compile the files control.o memory.o and freem.o without > optimization and openmp support (-O0 -qsmp-noopt). > > regards, > axel. > > BS> homepage, it coredumps. The same example runs fine on an executable > BS> compiled with 3.9.1 version, with a similar Makefile. > > was that executable compiled on the same machine? > with the same compilers? recently? > > also it might be helpful, if you could create a stack trace > of the core dump, so we now in what subroutine it dies. > > BS> Pls. find attached the output of the cpmd-3.9.2 run. > BS> > BS> We appreciate any help in this regard, > BS> > BS> Many Thanks! > BS> Bala > BS> > From ara_1357_2416 at yahoo.com Sat Aug 13 08:59:07 2005 From: ara_1357_2416 at yahoo.com (Younes Ansari) Date: Fri, 12 Aug 2005 23:59:07 -0700 (PDT) Subject: [CPMD-list] getting the output files just at the end Message-ID: <20050813065907.18055.qmail@web50907.mail.yahoo.com> Dear cpmd-list: I know that CPMD writes the output files after each time step. but I want to know if there would be a way to force cpmd writes the files just at the end of the running of the program. I have tested that the compress write wont solve the problem.updating the files at each step will take lots of time and it is time consuming. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050812/c3c41dbf/attachment.html From bala at jncasr.ac.in Sat Aug 13 10:28:35 2005 From: bala at jncasr.ac.in (Balasubramanian Sundaram) Date: Sat, 13 Aug 2005 13:58:35 +0530 Subject: [CPMD-list] Coredump on SP4 In-Reply-To: References: <42FC8B2A.30608@jncasr.ac.in> Message-ID: <42FDAF33.8070204@jncasr.ac.in> Dear Ari and Axel, Thanks for your suggestions. I used Axel's suggestion of configuring a Makefile with the "IBM-PWR5-AIX-ESSL-MPI-SMP" keyword, changed "pwr5" to "pwr4" everywhere, compiled the whole thing and ran it and it produced an output which is attached with this mail as file.out_02_392. It did not coredump. I now removed the -DPOINTER8 in CPPFLAGS, and it produced a core. I then recompiled control.f, memory.f and freem.f with the command: mpxlf_r -c -O0 -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 -qsmp=noopt control.f freem.f memory.f and then did a "make" and ran it and it too produced a core. The final Makefile is: ============================================================================== #--------------- Default Configuration for IBM-PWR5-AIX-ESSL-MPI-SMP --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 -qsmp=omp LFLAGS = -q64 -L/usr/local/lib -llapack64 -lesslsmp \ -qarch=pwr4 -qsmp=omp $(QMMM_LIBS) CFLAGS = -q64 -qarch=pwr4 CPP = /usr/ccs/lib/cpp -P CPPFLAGS = -D__IBM -DLAPACK -DFFT_ESSL -DMALLOC8 -DPARALLEL -DMP_LIBRARY=__MPI NOOPT_FLAG = -O0 -qsmp=noopt CC = xlc_r -c -O3 -qstrict FC = mpxlf_r -c -O3 -qstrict LD = mpxlf_r -O3 -qstrict AR = /usr/bin/ar ruv #---------------------------------------------------------------------------- =============================================================================== dbx gives: [using memory image in core] warning: Unable to access address 0x9001000a004c504 from core pthdb_session.c, 510: 1 PTHDB_CALLBACK (callback failed) pthreaded.c, 1778: PTHDB_CALLBACK (callback failed) Segmentation fault in memory at line 75 in file "/home/others/jncasr/NEW/CPMD-3.9.2/SOURCE/memory.f" 75 XM(0)=ZFIRST --------------- Thanks again, Regards, Bala Ari P Seitsonen wrote: > > Dear Bala, > > In addition to Axel's suggestions, have you removed the option > '-DPOINTER8' from the 'Makefile'? > > Greetings from Lausanne, > > apsi > > -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- > > Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ > CNRS & IMPMC, Universit? Pierre et Marie Curie > 4 place Jussieu, case 115 / F-75252 Paris > Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 > > On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: > >> Hello List Members, >> >> We are facing a difficulty in running cpmd on an IBM SP4. Here are >> the details: >> >> Version: CPMD-3.9.2 (May 14th version on cpmd.org site) >> >> The code compiles fine with the following header of the Makefile. >> #--------------- Default Configuration for IBM-SP4-SMP-64 --------------- >> SRC = . >> DEST = . >> BIN = . >> #QMMM_FLAGS = -D__QMECHCOUPL >> #QMMM_LIBS = -L. -lmm >> FFLAGS = -q64 -qtune=pwr4 -qarch=pwr4 -qsmp=omp >> LFLAGS = -L/usr/local/lib -llapack64 -lesslsmp -llapack64 -q64 \ >> -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr4 >> -qsmp=omp $(QMMM_LIBS) >> CFLAGS = $(FFLAGS) >> CPP = /usr/ccs/lib/cpp -P >> CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL=PARALLEL -DFFT_ESSL >> -DMP_LIBRARY=__MPI >> NOOPT_FLAG = >> CC = xlc_r -c >> FC = mpxlf_r -c >> LD = mpxlf_r >> AR = /usr/bin/ar >> #---------------------------------------------------------------------------- >> >> >> However, when we run the Si64-10ryd example (inp.wf) downloaded from >> Axel's homepage, it coredumps. The same example runs fine on an >> executable compiled with 3.9.1 version, with a similar Makefile. >> >> Pls. find attached the output of the cpmd-3.9.2 run. >> >> We appreciate any help in this regard, >> >> Many Thanks! >> Bala > > >> > > ------------------------------------------------------------------------ > > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: file.out_02_392 Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050813/f34500b9/attachment.cc From ari.p.seitsonen at iki.fi Sat Aug 13 11:26:30 2005 From: ari.p.seitsonen at iki.fi (Ari P Seitsonen) Date: Sat, 13 Aug 2005 11:26:30 +0200 (CEST) Subject: [CPMD-list] compile cpmd2cube In-Reply-To: <200508120655.IAA08390@internet-fence.zurich.ihost.com> References: <200508120655.IAA08390@internet-fence.zurich.ihost.com> Message-ID: Dear Jaime, Did you already manage to compile 'cpmd2cube'? Basically you can take almost any system and just adapt 'FC', 'FFLAGS' and 'FLAGS' to your system. In normal compilation you probably don't need any extra libraries. I'd try with 'FC = your-fortran-compiler', 'FFLAGS = -O' (or +O, if it's the HP convention) and 'FLAGS = [empty]'. Greetings, apsi -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ CNRS & IMPMC, Universit? Pierre et Marie Curie 4 place Jussieu, case 115 / F-75252 Paris Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 On Tue, 9 Aug 2005, Jaime Sanchez wrote: > Dear CPMD users, > > > > I want to compile cpmd2cube for the configuration HPUX-IA64 but this > configuration is not available when I type ./Configure. > > > > Anyone can help me with this? > > > > Thanks > > Jaime > > From ari.p.seitsonen at iki.fi Sat Aug 13 11:29:22 2005 From: ari.p.seitsonen at iki.fi (Ari P Seitsonen) Date: Sat, 13 Aug 2005 11:29:22 +0200 (CEST) Subject: [CPMD-list] Coredump on SP4 In-Reply-To: <42FDAF33.8070204@jncasr.ac.in> References: <42FC8B2A.30608@jncasr.ac.in> <42FDAF33.8070204@jncasr.ac.in> Message-ID: Dear Bala, Sorry, I meant that you _should_ ave '-DPOINTER8' always when you use '-q64'; it's so in the original architecture files but for some reason you apparently had removed it. Greetings, apsi PS Short explanation: '-q64' turns on 64-bit addresses, and CPMD stores the addresses in INTEGER variables; in 64-bit mode they have to be INTEGER*8, that's why the pre-processor flag '-DPOINTER8' in that case -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ CNRS & IMPMC, Universit? Pierre et Marie Curie 4 place Jussieu, case 115 / F-75252 Paris Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 On Sat, 13 Aug 2005, Balasubramanian Sundaram wrote: > Dear Ari and Axel, > Thanks for your suggestions. I used Axel's suggestion of configuring a > Makefile with the "IBM-PWR5-AIX-ESSL-MPI-SMP" keyword, changed "pwr5" to > "pwr4" everywhere, compiled the whole thing and ran it and it produced an > output which is attached with this mail as file.out_02_392. It did not > coredump. > > I now removed the -DPOINTER8 in CPPFLAGS, and it produced a core. > > I then recompiled control.f, memory.f and freem.f with the command: > mpxlf_r -c -O0 -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 -qsmp=noopt > control.f freem.f memory.f > > and then did a "make" and ran it and it too produced a core. > > The final Makefile is: > ============================================================================== > #--------------- Default Configuration for IBM-PWR5-AIX-ESSL-MPI-SMP > --------------- > SRC = . > DEST = . > BIN = . > #QMMM_FLAGS = -D__QMECHCOUPL > #QMMM_LIBS = -L. -lmm > FFLAGS = -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 -qsmp=omp > LFLAGS = -q64 -L/usr/local/lib -llapack64 -lesslsmp \ > -qarch=pwr4 -qsmp=omp $(QMMM_LIBS) > CFLAGS = -q64 -qarch=pwr4 > CPP = /usr/ccs/lib/cpp -P > CPPFLAGS = -D__IBM -DLAPACK -DFFT_ESSL -DMALLOC8 -DPARALLEL > -DMP_LIBRARY=__MPI > NOOPT_FLAG = -O0 -qsmp=noopt > CC = xlc_r -c -O3 -qstrict > FC = mpxlf_r -c -O3 -qstrict > LD = mpxlf_r -O3 -qstrict > AR = /usr/bin/ar ruv > #---------------------------------------------------------------------------- > =============================================================================== > > dbx gives: > [using memory image in core] > warning: Unable to access address 0x9001000a004c504 from core > pthdb_session.c, 510: 1 PTHDB_CALLBACK (callback failed) > pthreaded.c, 1778: PTHDB_CALLBACK (callback failed) > > Segmentation fault in memory at line 75 in file > "/home/others/jncasr/NEW/CPMD-3.9.2/SOURCE/memory.f" > 75 XM(0)=ZFIRST > --------------- > > Thanks again, > > Regards, > Bala > > > > Ari P Seitsonen wrote: >> >> Dear Bala, >> >> In addition to Axel's suggestions, have you removed the option >> '-DPOINTER8' from the 'Makefile'? >> >> Greetings from Lausanne, >> >> apsi >> >> >> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- >> Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ >> CNRS & IMPMC, Universit? Pierre et Marie Curie >> 4 place Jussieu, case 115 / F-75252 Paris >> Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 >> >> On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: >> >>> Hello List Members, >>> >>> We are facing a difficulty in running cpmd on an IBM SP4. Here are the >>> details: >>> >>> Version: CPMD-3.9.2 (May 14th version on cpmd.org site) >>> >>> The code compiles fine with the following header of the Makefile. >>> #--------------- Default Configuration for IBM-SP4-SMP-64 --------------- >>> SRC = . >>> DEST = . >>> BIN = . >>> #QMMM_FLAGS = -D__QMECHCOUPL >>> #QMMM_LIBS = -L. -lmm >>> FFLAGS = -q64 -qtune=pwr4 -qarch=pwr4 -qsmp=omp >>> LFLAGS = -L/usr/local/lib -llapack64 -lesslsmp -llapack64 -q64 \ >>> -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr4 >>> -qsmp=omp $(QMMM_LIBS) >>> CFLAGS = $(FFLAGS) >>> CPP = /usr/ccs/lib/cpp -P >>> CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL=PARALLEL -DFFT_ESSL >>> -DMP_LIBRARY=__MPI >>> NOOPT_FLAG = >>> CC = xlc_r -c >>> FC = mpxlf_r -c >>> LD = mpxlf_r >>> AR = /usr/bin/ar >>> #---------------------------------------------------------------------------- >>> >>> However, when we run the Si64-10ryd example (inp.wf) downloaded from >>> Axel's homepage, it coredumps. The same example runs fine on an executable >>> compiled with 3.9.1 version, with a similar Makefile. >>> >>> Pls. find attached the output of the cpmd-3.9.2 run. >>> >>> We appreciate any help in this regard, >>> >>> Many Thanks! >>> Bala >> >> >>> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> CPMD-list mailing list >> CPMD-list at cpmd.org >> http://cpmd.org/mailman/listinfo/cpmd-list > From bala at jncasr.ac.in Sat Aug 13 11:30:44 2005 From: bala at jncasr.ac.in (Balasubramanian Sundaram) Date: Sat, 13 Aug 2005 15:00:44 +0530 Subject: [CPMD-list] Coredump on SP4 In-Reply-To: References: <42FC8B2A.30608@jncasr.ac.in> <42FDAF33.8070204@jncasr.ac.in> Message-ID: <42FDBDC4.7010009@jncasr.ac.in> dear ari, Thanks. I did have it, in one of the early trials. (it too produced a core). best, bala Ari P Seitsonen wrote: > > Dear Bala, > > Sorry, I meant that you _should_ ave '-DPOINTER8' always when you use > '-q64'; it's so in the original architecture files but for some reason > you apparently had removed it. > > Greetings, > > apsi > > PS Short explanation: '-q64' turns on 64-bit addresses, and CPMD stores > the addresses in INTEGER variables; in 64-bit mode they have to be > INTEGER*8, that's why the pre-processor flag '-DPOINTER8' in that case > > -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- > > Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ > CNRS & IMPMC, Universit? Pierre et Marie Curie > 4 place Jussieu, case 115 / F-75252 Paris > Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 > > On Sat, 13 Aug 2005, Balasubramanian Sundaram wrote: > >> Dear Ari and Axel, >> Thanks for your suggestions. I used Axel's suggestion of >> configuring a Makefile with the "IBM-PWR5-AIX-ESSL-MPI-SMP" keyword, >> changed "pwr5" to "pwr4" everywhere, compiled the whole thing and ran >> it and it produced an output which is attached with this mail as >> file.out_02_392. It did not coredump. >> >> I now removed the -DPOINTER8 in CPPFLAGS, and it produced a core. >> >> I then recompiled control.f, memory.f and freem.f with the command: >> mpxlf_r -c -O0 -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 >> -qsmp=noopt control.f freem.f memory.f >> >> and then did a "make" and ran it and it too produced a core. >> >> The final Makefile is: >> ============================================================================== >> >> #--------------- Default Configuration for IBM-PWR5-AIX-ESSL-MPI-SMP >> --------------- >> SRC = . >> DEST = . >> BIN = . >> #QMMM_FLAGS = -D__QMECHCOUPL >> #QMMM_LIBS = -L. -lmm >> FFLAGS = -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 -qsmp=omp >> LFLAGS = -q64 -L/usr/local/lib -llapack64 -lesslsmp \ >> -qarch=pwr4 -qsmp=omp $(QMMM_LIBS) >> CFLAGS = -q64 -qarch=pwr4 >> CPP = /usr/ccs/lib/cpp -P >> CPPFLAGS = -D__IBM -DLAPACK -DFFT_ESSL -DMALLOC8 -DPARALLEL >> -DMP_LIBRARY=__MPI >> NOOPT_FLAG = -O0 -qsmp=noopt >> CC = xlc_r -c -O3 -qstrict >> FC = mpxlf_r -c -O3 -qstrict >> LD = mpxlf_r -O3 -qstrict >> AR = /usr/bin/ar ruv >> #---------------------------------------------------------------------------- >> >> =============================================================================== >> >> >> dbx gives: >> [using memory image in core] >> warning: Unable to access address 0x9001000a004c504 from core >> pthdb_session.c, 510: 1 PTHDB_CALLBACK (callback failed) >> pthreaded.c, 1778: PTHDB_CALLBACK (callback failed) >> >> Segmentation fault in memory at line 75 in file >> "/home/others/jncasr/NEW/CPMD-3.9.2/SOURCE/memory.f" >> 75 XM(0)=ZFIRST >> --------------- >> >> Thanks again, >> >> Regards, >> Bala >> >> >> >> Ari P Seitsonen wrote: >> >>> >>> Dear Bala, >>> >>> In addition to Axel's suggestions, have you removed the option >>> '-DPOINTER8' from the 'Makefile'? >>> >>> Greetings from Lausanne, >>> >>> apsi >>> >>> >>> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- >>> >>> Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ >>> CNRS & IMPMC, Universit? Pierre et Marie Curie >>> 4 place Jussieu, case 115 / F-75252 Paris >>> Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 >>> >>> On Fri, 12 Aug 2005, Balasubramanian Sundaram wrote: >>> >>>> Hello List Members, >>>> >>>> We are facing a difficulty in running cpmd on an IBM SP4. Here are >>>> the details: >>>> >>>> Version: CPMD-3.9.2 (May 14th version on cpmd.org site) >>>> >>>> The code compiles fine with the following header of the Makefile. >>>> #--------------- Default Configuration for IBM-SP4-SMP-64 >>>> --------------- >>>> SRC = . >>>> DEST = . >>>> BIN = . >>>> #QMMM_FLAGS = -D__QMECHCOUPL >>>> #QMMM_LIBS = -L. -lmm >>>> FFLAGS = -q64 -qtune=pwr4 -qarch=pwr4 -qsmp=omp >>>> LFLAGS = -L/usr/local/lib -llapack64 -lesslsmp -llapack64 -q64 \ >>>> -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr4 >>>> -qsmp=omp $(QMMM_LIBS) >>>> CFLAGS = $(FFLAGS) >>>> CPP = /usr/ccs/lib/cpp -P >>>> CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL=PARALLEL -DFFT_ESSL >>>> -DMP_LIBRARY=__MPI >>>> NOOPT_FLAG = >>>> CC = xlc_r -c >>>> FC = mpxlf_r -c >>>> LD = mpxlf_r >>>> AR = /usr/bin/ar >>>> #---------------------------------------------------------------------------- >>>> >>>> However, when we run the Si64-10ryd example (inp.wf) downloaded from >>>> Axel's homepage, it coredumps. The same example runs fine on an >>>> executable compiled with 3.9.1 version, with a similar Makefile. >>>> >>>> Pls. find attached the output of the cpmd-3.9.2 run. >>>> >>>> We appreciate any help in this regard, >>>> >>>> Many Thanks! >>>> Bala >>> >>> >>> >>>> >>> >>> ------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> CPMD-list mailing list >>> CPMD-list at cpmd.org >>> http://cpmd.org/mailman/listinfo/cpmd-list > > >> From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Sat Aug 13 12:23:24 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Sat, 13 Aug 2005 12:23:24 +0200 (CEST) Subject: [CPMD-list] Coredump on SP4 In-Reply-To: <42FDAF33.8070204@jncasr.ac.in> Message-ID: On Sat, 13 Aug 2005, Balasubramanian Sundaram wrote: dear bala, BS> Dear Ari and Axel, BS> Thanks for your suggestions. I used Axel's suggestion of configuring a BS> Makefile with the "IBM-PWR5-AIX-ESSL-MPI-SMP" keyword, changed "pwr5" to "pwr4" BS> everywhere, compiled the whole thing and ran it and it produced an output which is BS> attached with this mail as file.out_02_392. It did not coredump. hmmm, i've seen that one before. i think it was related to OpenMP. please try the IBM-PWR5-AIX-ESSL-MPI (withOUT SMP) config and use pwr4 instead of pwr5. i noticed recently that this bulk silicon example can behave a little strange sometimes. you many want to try a different input, too. 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 bala at jncasr.ac.in Sat Aug 13 13:12:42 2005 From: bala at jncasr.ac.in (Balasubramanian Sundaram) Date: Sat, 13 Aug 2005 16:42:42 +0530 Subject: [CPMD-list] Coredump on SP4 In-Reply-To: References: Message-ID: <42FDD5AA.4090408@jncasr.ac.in> Dear Axel, Thanks for all your suggestions. The executable that was compiled without SMP is able to run the Si64-10ryd (inp.wf) as well as other input files which are specific to our research work. Thanks a million! If you require my help to further investigate the issue(s) with respect to the SMP version on this platform, please let me know. I would require a lot of handholding though... Regards, Bala ===========Final Makefile that worked========================= #--------------- Default Configuration for IBM-PWR5-AIX-ESSL-MPI --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -q64 -qmaxmem=32768 -qtune=pwr4 -qarch=pwr4 LFLAGS = -q64 -L/usr/local/lib -llapack64 -lessl \ -qarch=pwr4 $(QMMM_LIBS) CFLAGS = -q64 -qarch=pwr4 CPP = /usr/ccs/lib/cpp -P CPPFLAGS = -D__IBM -DLAPACK -DFFT_ESSL -DMALLOC8 -DPOINTER8 -DPARALLEL=PARALLEL -DMP_LIBRARY=__MPI NOOPT_FLAG = -O0 CC = xlc_r -c -O3 -qstrict FC = mpxlf_r -c -O3 -qstrict LD = mpxlf_r -O3 -qstrict AR = /usr/bin/ar ruv #---------------------------------------------------------------------------- Axel Kohlmeyer wrote: > On Sat, 13 Aug 2005, Balasubramanian Sundaram wrote: > > dear bala, > > BS> Dear Ari and Axel, > BS> Thanks for your suggestions. I used Axel's suggestion of configuring a > BS> Makefile with the "IBM-PWR5-AIX-ESSL-MPI-SMP" keyword, changed "pwr5" to "pwr4" > BS> everywhere, compiled the whole thing and ran it and it produced an output which is > BS> attached with this mail as file.out_02_392. It did not coredump. > > hmmm, i've seen that one before. i think it was related to OpenMP. > please try the IBM-PWR5-AIX-ESSL-MPI (withOUT SMP) config and > use pwr4 instead of pwr5. > > i noticed recently that this bulk silicon example can behave > a little strange sometimes. you many want to try a different > input, too. > > regards, > axel. > > [...] > From lyuan at Princeton.EDU Mon Aug 15 02:41:17 2005 From: lyuan at Princeton.EDU (Lan-Feng Yuan (lyuan@Princeton.EDU)) Date: Sun, 14 Aug 2005 20:41:17 -0400 Subject: [CPMD-list] thermostat "frequency": mu or omega? Message-ID: Hi, dear CPMD users, We know the vibrational frequencies of H2O molecule are ~ 3800 and 1600 cm-1, then we want to select a thermostat frequency in between, say 2000 cm-1. Then, should I write "NOSE IONS 300. 2000." or "NOSE IONS 300. 1256."? In the latter, I mean the angular frequency omega = 2*pi*mu = 1256 cm-1. Why am I asking this? Because when I try to do path integral calculations, I always get information like (number of replicas P=16, T=300 K): THERMOSTATS: CHANGE IONIC NOSE FREQUENCIES IN PATH INTEGRAL CASE FOR PRIMITIVE PROPAGATOR CHARACTERISTIC FREQUENCY FOR IP=1 : 5234.37 CM**-1 CHARACTERISTIC FREQUENCY FOR IP>1 : 5234.37 CM**-1 No matter what frequency I select in the input, it is changed to a certain value by the program. The reason is that the coupling between path integral beads is perfectly harmonic and very hard to get thermolized, so CPMD switches the thermostat frequency to the value corresponding to this coupling. That's a good idea. But it seems to me that this value shown is omega rather than mu. This harmonic coupling has the form of 1/2*m*P*(k_B*T/hbar)^2*(r_i-r_i+1)^2, therefore the so-called omega_P=sqrt(P)*k_B*T/hbar. For P=16 and T=300 K, I get this omega_P=5234 cm-1. It's noteworthy that its frequency mu is ~ 834 cm-1, so it's actually lower than the lowest vibrational mode of water (1600 cm-1), not higher than the highest (3800 cm-1)! First, I hope this is not a mistake ... Second, since the "frequency" here is omega, does that mean I should input omega also in other cases? Thank you very much! Best regards, Lan-Feng ---------------------------- Dr. Lan-Feng Yuan, Postdoctoral Associate, Department of Chemistry, Princeton University, Princeton, NJ 08544, USA Office Phone: 1-609-258-1834 Fax: 1-609-258-6746 ---------------------------- From hutter at pci.unizh.ch Sun Aug 14 18:11:23 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Sun, 14 Aug 2005 18:11:23 +0200 (MEST) Subject: [CPMD-list] ROKS and TDDFT of spin polarized systems In-Reply-To: <1123858090.42fcb6aa4ebd9@impmail.kcl.ac.uk> References: <1123858090.42fcb6aa4ebd9@impmail.kcl.ac.uk> Message-ID: Hi > Hello everyone, > > Here I ask another question. I am studying the isomerization of a > chromophore of a photoreceptor protein. I am doing that by calculating > potential energy surfaces of the ground and first excited singlet state along > the isomerization coordinate, the torsional angle around a double bond. I am > calculating the excited states with ROKS and TDDFT. At the twisted angle of 90 > degrees in the ground state the double bond breaks and a spin unrestricted > (LSD) calculation is necessary. My question is: how can I calculate the excited > state of that structure? > > - I tried to do the calculation with TDDFT but only a mixed state was obtained. > Is there the spin-unrestricted TDDFT implemented on CPMD? Yes, there is. Use the LSD keyword and ask for MIXED STATES in the TDDFT section. > > - ROKS is a spin restricted method, but could I somehow do different > calculatiosn by transfering the electron at HOMO with spin-up and spin-down and > then do a linear combination to get the right spin? > I don't get what you want to achieve with this. ROKS gives you a singlet wavefunction in the excited state. It should also be able to handle a biradical reasonably (see the literature for the ROSS method). There have been several publications on double bond isomerisation using ROKS and TDDFT. There you will find all the problems people have encountered for this type of excited state reaction. regards Juerg Hutter > Thank you very much for any hints or comments. > > Regards, > > > Eneritz > > > -- > Eneritz Muguruza > eneritz.muguruza at kcl.ac.uk > > > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list > From hutter at pci.unizh.ch Sun Aug 14 18:06:08 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Sun, 14 Aug 2005 18:06:08 +0200 (MEST) Subject: [CPMD-list] getting the output files just at the end In-Reply-To: <20050813065907.18055.qmail@web50907.mail.yahoo.com> References: <20050813065907.18055.qmail@web50907.mail.yahoo.com> Message-ID: Hi > Dear cpmd-list: > I know that CPMD writes the output files after each time step. but I want to know if there would be a way to force cpmd writes the files just at the end of the running of the program. I have tested that the compress write wont solve the problem.updating the files at each step will take lots of time and it is time consuming. > Are you sure that this is a problem? My experience is that writing files is only a problem on extremly fast computers (with 100's of CPUs). But if you want you can try to increase the file buffer and decrease the frequency of the trajectory writing (see manual). regards Juerg Hutter > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com From eyvaz_isaev at yahoo.com Mon Aug 15 23:49:56 2005 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 15 Aug 2005 14:49:56 -0700 (PDT) Subject: [CPMD-list] Band gap and hybride EXC In-Reply-To: Message-ID: <20050815214956.55810.qmail@web60324.mail.yahoo.com> Dear users, Is there any recipie to calculate band gap for semiconductors, say for Si, using hybride EXC (ACM0, ACM1, etc.)? Should I first calculate optimized wavefunctions and then start for EXC calculations? Should wavefunction optimization be carried out using just the LDA functional? Any hint will be appreciated. Bests, Eyvaz. ____________________________________________________ Start your day with Yahoo! - make it your home page http://www.yahoo.com/r/hs From hutter at pci.unizh.ch Tue Aug 16 17:26:32 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Tue, 16 Aug 2005 17:26:32 +0200 (MEST) Subject: [CPMD-list] Band gap and hybride EXC In-Reply-To: <20050815214956.55810.qmail@web60324.mail.yahoo.com> References: <20050815214956.55810.qmail@web60324.mail.yahoo.com> Message-ID: Hi > Dear users, > > Is there any recipie to calculate band gap for > semiconductors, say for Si, using hybride EXC (ACM0, > ACM1, etc.)? Should I first calculate optimized > wavefunctions and then start for EXC calculations? > Should wavefunction optimization be carried out using > just the LDA functional? The hybrid functionals have to be applied in a self-consistent calculation as any other functionals. Starting with optimized wavefunctions (from another functional) might help to reduce the number of iterations needed but is not essential for the calculation otherwise. Some words of caution: - hybrid functionals are not implemented with k-points - there is a problem with a divergence for periodic boundary conditions. The standard way of dealing with this is by adding an integrable compensation charge. This is not implemented in CPMD. The only way to handle PBC at the moment is by using screened exchange. This will change the functional and has to be tested case by case. regards Juerg Hutter > > Any hint will be appreciated. > > Bests, > Eyvaz. > > > > ____________________________________________________ > Start your day with Yahoo! - make it your home page > http://www.yahoo.com/r/hs > > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list > From irg20 at chem.ubc.ca Tue Aug 16 19:06:34 2005 From: irg20 at chem.ubc.ca (Ian R. Greig) Date: Tue, 16 Aug 2005 10:06:34 -0700 (PDT) Subject: [CPMD-list] CPMD/Gromos compile on IBM-SP5 Message-ID: Dear CPMD users, I am trying to compile the CPMD/Gromos code of Roethlisberger et al. on an IBM power 5 550 machine. I am using the following Makefile (based on the IBM-SP4-SMP-64 Makefile for CPMD of Dr Curioni)... #---------------------------------------------------------------------------- # Makefile for cpmd.x (plane wave electronic calculation) # Configuration: IBM-SP3 # Creation of Makefile: Aug 15 2005 # on AIX cortex 3 5 00CC3C7E4C00 # Author: irg20 #---------------------------------------------------------------------------- # SHELL = /bin/sh # #--------------- Default Configuration for IBM-SP3 --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -O3 -qstrict -q64 -qmaxmem=-1 -qtune=pwr5 -qarch=pwr5 \ -qsmp=omp LFLAGS = -L/scratch_many/irg20/ibm/LAPACK/ -llapack -lessl -q64 \ -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 \ -bloadmap:loaderror.ian \ -qsmp=omp CFLAGS = -O3 -qstrict -qarch=pwr5 CPP = /usr/ccs/lib/cpp -P CPPFLAGS = -D__IBM -DLAPACK -DPARALLEL -DMP_LIBRARY=__MPI -DFFT_ESSL \ -DPOINTER8 -D__GROMOS CPPFLAGS_GROMOS = -DEWALD -DEWATCUT -DHAT_SHAPE -UNPACKED_GRID FFLAGS_GROMOS = -c -O3 -qarch=pwr5 -qdpc CC = cc FC = mpxlf -c -O3 -qstrict LD = mpxlf -O3 -qstrict AR = /usr/bin/ar #---------------------------------------------------------------------------- # # LIST OF FILES # etc... As suggested I have compiled control.f, memory.f and freem.f without the -qsmp -O3 flags. The files compile fine but at the loader stage I get the fllowing error message... ld: 0711-738 ERROR: Input file sysdepend.o: XCOFF32 object files are not allowed in 64-bit mode. make: 1254-004 The error code from the last command is 8. I was wondering whether anyone could provide me with any help if they have experience compiling on this system. Many thanks, Ian Greig From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Tue Aug 16 19:40:03 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Tue, 16 Aug 2005 19:40:03 +0200 (CEST) Subject: [CPMD-list] CPMD/Gromos compile on IBM-SP5 In-Reply-To: Message-ID: On Tue, 16 Aug 2005, Ian R. Greig wrote: IG> Dear CPMD users, ian, IG> I am trying to compile the CPMD/Gromos code of Roethlisberger et al. on an IG> IBM power 5 550 machine. [...] IG> FFLAGS = -O3 -qstrict -q64 -qmaxmem=-1 -qtune=pwr5 -qarch=pwr5 \ IG> -qsmp=omp IG> LFLAGS = -L/scratch_many/irg20/ibm/LAPACK/ -llapack -lessl -q64 \ IG> -bbinder:/usr/lib/bind -bmaxdata:32768000000 -qarch=pwr5 \ IG> -bloadmap:loaderror.ian \ IG> -qsmp=omp IG> CFLAGS = -O3 -qstrict -qarch=pwr5 [...] IG> As suggested I have compiled control.f, memory.f and freem.f without the IG> -qsmp -O3 flags. IG> IG> The files compile fine but at the loader stage I get the fllowing error IG> message... IG> IG> ld: 0711-738 ERROR: Input file sysdepend.o: IG> XCOFF32 object files are not allowed in 64-bit mode. IG> make: 1254-004 The error code from the last command is 8. IG> IG> I was wondering whether anyone could provide me with any help if they have IG> experience compiling on this system. even though, i have no experience compiling on this system , my guess is, that you are missing the -q64 flag in the CFLAGS definition. please try to add it, remove sysdepend.o and type make. good luck, axel. IG> Many thanks, IG> IG> Ian Greig IG> IG> IG> _______________________________________________ IG> CPMD-list mailing list IG> CPMD-list at cpmd.org IG> http://cpmd.org/mailman/listinfo/cpmd-list IG> IG> -- ======================================================================= 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 iivanov at vitae.cmm.upenn.edu Tue Aug 23 01:17:17 2005 From: iivanov at vitae.cmm.upenn.edu (Ivaylo Ivanov) Date: Mon, 22 Aug 2005 19:17:17 -0400 (EDT) Subject: [CPMD-list] CPMD on BlueGene Message-ID: <3345.132.239.16.215.1124752637.squirrel@cmm.upenn.edu> Dear CPMD users, I would like to find out if anyone has tried/succeeded in porting the CPMD code to the IBM BlueGene architecture. Please, let me know what are the necessary changes in the code as it appears that simple recompilation with new compiler options/Makefile doesn't work (The code compiles but then produces segmentation faults). SDSC has installed a one rack BG machine here and I would very much like to see if it is useful for CPMD. Before trying to debug I wanted to get expert opinion and hence this message. Thank you very much in advance. Sincerely, Ivaylo Ivanov -- Ivaylo Ivanov Center for Molecular Modeling | University of Pennsylvania 231 S.34 Street, | Philadelphia,PA 19104-6323 tel: (215) 573 8697 | fax: (215) 573 6233 e-mail: iivanov at cmm.chem.upenn.edu From cur at zurich.ibm.com Tue Aug 23 09:51:12 2005 From: cur at zurich.ibm.com (Alessandro Curioni) Date: Tue, 23 Aug 2005 09:51:12 +0200 Subject: Fw: [CPMD-list] CPMD on BlueGene Message-ID: Alessandro Curioni/Zurich/IBM 08/23/2005 09:31 AM To "Ivaylo Ivanov" cc Subject Re: [CPMD-list] CPMD on BlueGene Dear Ivaylo, yes - CPMD has been succesfully ported to BG/L - and will be distributed in the next version (very soon). Please have a look at http://www.sc-conference.org/sc2004/schedule/pdfs/pap220.pdf or http://www3.interscience.wiley.com/cgi-bin/fulltext/110577400/PDFSTART . In order to take advantage of the BG/L hardware - several modifications were required - you can get 3.9.2 code running as is - but it will be ~4 times slower. I am in contact with giri chukkapalli of USDC - he will get the beta version and install it there - so I suggest you contact him. 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 "Ivaylo Ivanov" Sent by: cpmd-list-bounces at cpmd.org 08/23/2005 01:17 AM To cpmd-list at cpmd.org cc Subject [CPMD-list] CPMD on BlueGene Dear CPMD users, I would like to find out if anyone has tried/succeeded in porting the CPMD code to the IBM BlueGene architecture. Please, let me know what are the necessary changes in the code as it appears that simple recompilation with new compiler options/Makefile doesn't work (The code compiles but then produces segmentation faults). SDSC has installed a one rack BG machine here and I would very much like to see if it is useful for CPMD. Before trying to debug I wanted to get expert opinion and hence this message. Thank you very much in advance. Sincerely, Ivaylo Ivanov -- Ivaylo Ivanov Center for Molecular Modeling | University of Pennsylvania 231 S.34 Street, | Philadelphia,PA 19104-6323 tel: (215) 573 8697 | fax: (215) 573 6233 e-mail: iivanov at cmm.chem.upenn.edu _______________________________________________ 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/20050823/ac41a0af/attachment.html From mattwagner at hotmail.com Tue Aug 23 16:23:03 2005 From: mattwagner at hotmail.com (Matt wagner) Date: Tue, 23 Aug 2005 14:23:03 +0000 Subject: [CPMD-list] 143 steps too long? Message-ID: I am doing single point calculation for density minimization on BH4, cation. Everything worked fined, but took 143 steps to complete. Is this too long? Incorrect? I first used ODIIS and it took more than 1000 steps for wavefunction optimization. When I used PCG Minimizer, that is when I got it down to 143 steps. This was all done using the Vanderbelt psuedopotential. The manual says that steps larger than 100 is to be considered slow. Does slow mean inaccurate in this case? Any suggestions would be greatly appreciated. Below is my input file: Matt Wagner Cleveland State University Chemistry Dept. mattwagner at hotmail.com &CPMD WAVEFUNCTION OPTIMIZATION PCG MINIMIZE TIMESTEP 20 &END &DFT FUNCTIONAL PBE &END &SYSTEM SYMMETRY 0 CHARGE 1.0 CELL 40.0000000 1.0000000 1.0000000 0.0 0.0 0.0 CUTOFF 80 &END &ATOMS *H_MT_PBE.psp KLEINMAN-BYLANDER LMAX=S 15 10.2206634 20.0283459 17.1559610 10.1261771 16.8725020 17.1559610 14.6615216 19.6881951 13.7922472 21.9369701 21.3700520 18.4409753 22.6928608 16.6268375 21.2566684 20.3118049 21.1999766 23.9211833 20.8220312 18.1008245 24.9794303 16.7780157 17.9118518 24.0912587 25.9053965 19.3480442 19.8960650 26.1510610 23.3731625 21.3511547 24.6959713 19.6692978 25.4707593 27.9651988 16.6646321 25.4518620 29.4958776 20.3307022 26.2077528 29.8738229 20.1228322 23.0141144 28.0029933 22.7684499 24.5636904 *O_MT_PBE.psp KLEINMAN-BYLANDER LMAX=P 3 18.9889961 20.7653394 14.0568090 24.4881013 22.5983745 21.3133602 26.7179791 16.7024266 24.1101560 *C_MT_PBE.psp KLEINMAN-BYLANDER LMAX=P 9 13.7166581 18.3464890 17.3071392 17.8929545 19.7826814 15.8709468 18.8189207 18.9512016 18.2331054 17.0614747 18.5165644 20.1039350 22.2393264 18.6299480 21.2377711 20.3307022 19.2157633 23.3353679 24.7904576 19.9527568 21.5212301 26.0943692 19.2913524 24.0156696 28.5321169 20.7842367 24.4314095 *N_MT_PBE.psp KLEINMAN-BYLANDER LMAX=P 5 14.5292407 17.9874409 19.6315033 11.1277323 18.4220781 16.8158102 15.3229260 19.2157633 15.5307959 21.3322575 19.5748115 18.8000234 17.8929545 18.2142081 22.5416827 &END From pis_diez at quimica.unlp.edu.ar Tue Aug 23 20:20:08 2005 From: pis_diez at quimica.unlp.edu.ar (Reinaldo Pis Diez) Date: Tue, 23 Aug 2005 15:20:08 -0300 Subject: [CPMD-list] Bad results during the heating phase of a md run Message-ID: <430B68D8.3050203@quimica.unlp.edu.ar> Dear folks, I'm trying to reproduce some older results on sodium clusters (R?thlisberger and Andreoni, JCP 94 (1991) 8129) following a different strategy. I want to let the clusters evolve at a given temperature for a long period saving the coordinates every several ps and then optimize each previously saved geometry. I'm using version 3.9.2 of CPMD, Goedecker pseudos for sodium with Zion = 9 and PBE xc functional. I've made some previous tests with the sodium dimer and found that 60 Ry for PW cutoff are ok to reproduce the bond distance, for example. Then, I turn to the sodium hexamer. I use a simulated annealing procedure to optimize the wavefunction (na6_wfn.inp and na6_wfn.out files) and it seems to be alright. Then, the hexamer is allowed to heat up to 600 K. The lattice of the (cubic) cell was 15 A and the PW cutoff was 60 Ry as said above. The md run diverges from the second time step. I increased the lattice cell and the PW cutoff to 25 A and 70 Ry, respectively, to see where the problem is but I got the same results (na6_md_ini.inp and na6_md_ini.out files). I can see from the output that EKINC is going large at the very beginning of the run but I don't know if this could be a clue and how to handle it to get the simulation running. Any help will be very appreciated. The input and output files are appended. Sorry for the long mail but I cannot figure out how to explain my problem in few words. Thanks in advance. Regards, Reinaldo -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: na6_md_ini.inp Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050823/c601a078/attachment.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: na6_md_ini.out Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050823/c601a078/attachment-0001.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: na6_wfn.inp Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050823/c601a078/attachment-0002.cc -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: na6_wfn.out Url: http://cpmd.org/pipermail/cpmd-list/attachments/20050823/c601a078/attachment-0003.cc From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Wed Aug 24 13:08:17 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Wed, 24 Aug 2005 13:08:17 +0200 (CEST) Subject: [CPMD-list] Bad results during the heating phase of a md run In-Reply-To: <430B68D8.3050203@quimica.unlp.edu.ar> Message-ID: On Tue, 23 Aug 2005, Reinaldo Pis Diez wrote: reinaldo, [...] RPD> Zion = 9 and PBE xc functional. I've made some previous tests with the RPD> sodium dimer and found that 60 Ry for PW cutoff are ok to reproduce RPD> the bond distance, for example. Then, I turn to the sodium hexamer. I RPD> use a simulated annealing procedure to optimize the wavefunction RPD> (na6_wfn.inp and na6_wfn.out files) and it seems to be alright. Then, RPD> the hexamer is allowed to heat up to 600 K. The lattice of the (cubic) RPD> cell was 15 A and the PW cutoff was 60 Ry as said above. The md run RPD> diverges from the second time step. I increased the lattice cell and this is because you did not read the coordinates (and velocities) of your atoms from the restart file, so the wavefunction did not match the positions of your coordinates. note that you have a significant temperature of the ions (TEMPP) in your annealing. (btw. is there are reason that you use that instead of optimizing the wavefunction with ODIIS of PCG?). so you either have to add COORDINATES (and VELOCITIES) to the RESTART line in your MD or add the following to the annealing input in the &ATOMS section: CONSTRAINTS FIX ALL ATOMS END CONSTRAINTS RPD> the PW cutoff to 25 A and 70 Ry, respectively, to see where the RPD> problem is but I got the same results (na6_md_ini.inp and RPD> na6_md_ini.out files). I can see from the output that EKINC is going RPD> large at the very beginning of the run but I don't know if this could RPD> be a clue and how to handle it to get the simulation running. Any RPD> help will be very appreciated. The input and output files are appended. RPD> Sorry for the long mail but I cannot figure out how to explain my RPD> problem in few words. no problem. best regards, axel. RPD> Thanks in advance. RPD> Regards, RPD> RPD> Reinaldo RPD> -- ======================================================================= 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 hutter at pci.unizh.ch Wed Aug 24 13:32:13 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Wed, 24 Aug 2005 13:32:13 +0200 (MEST) Subject: [CPMD-list] Bad results during the heating phase of a md run In-Reply-To: <430B68D8.3050203@quimica.unlp.edu.ar> References: <430B68D8.3050203@quimica.unlp.edu.ar> Message-ID: Hi there seems to be several problems with your simulation. Do you start the simulation from optimized wavefunctions? In your first simulation you have very large initial forces (up to 15!). This will cause a large initial step and the wavefunction (even when converged) will not be able to follow. You should first relax the geometry either using Geometry optimization or MD with annealing and then start an MD and test for the optimal EMASS and timestep. regards Juerg Hutter ---------------------------------------------------------- Juerg Hutter Phone : ++41 44 635 4491 Physical Chemistry Institute FAX : ++41 44 635 6838 University of Zurich E-mail: hutter at pci.unizh.ch Winterthurerstrasse 190 CH-8057 Zurich, Switzerland ---------------------------------------------------------- On Tue, 23 Aug 2005, Reinaldo Pis Diez wrote: > Dear folks, > > I'm trying to reproduce some older results on sodium clusters > (R?thlisberger and Andreoni, JCP 94 (1991) 8129) following a different > strategy. I want to let the clusters evolve at a given temperature for > a long period saving the coordinates every several ps and then > optimize each previously saved geometry. > I'm using version 3.9.2 of CPMD, Goedecker pseudos for sodium with > Zion = 9 and PBE xc functional. I've made some previous tests with the > sodium dimer and found that 60 Ry for PW cutoff are ok to reproduce > the bond distance, for example. Then, I turn to the sodium hexamer. I > use a simulated annealing procedure to optimize the wavefunction > (na6_wfn.inp and na6_wfn.out files) and it seems to be alright. Then, > the hexamer is allowed to heat up to 600 K. The lattice of the (cubic) > cell was 15 A and the PW cutoff was 60 Ry as said above. The md run > diverges from the second time step. I increased the lattice cell and > the PW cutoff to 25 A and 70 Ry, respectively, to see where the > problem is but I got the same results (na6_md_ini.inp and > na6_md_ini.out files). I can see from the output that EKINC is going > large at the very beginning of the run but I don't know if this could > be a clue and how to handle it to get the simulation running. Any > help will be very appreciated. The input and output files are appended. > Sorry for the long mail but I cannot figure out how to explain my > problem in few words. > Thanks in advance. > Regards, > > Reinaldo > From hutter at pci.unizh.ch Wed Aug 24 13:37:15 2005 From: hutter at pci.unizh.ch (Juerg Hutter) Date: Wed, 24 Aug 2005 13:37:15 +0200 (MEST) Subject: [CPMD-list] 143 steps too long? In-Reply-To: References: Message-ID: Hi > I am doing single point calculation for density minimization on BH4, > cation. Everything worked fined, but took 143 steps to complete. Is this > too long? Incorrect? I first used ODIIS and it took more than 1000 steps > for wavefunction optimization. When I used PCG Minimizer, that is when I > got it down to 143 steps. This was all done using the Vanderbelt > psuedopotential. The manual says that steps larger than 100 is to be > considered slow. Does slow mean inaccurate in this case? Any suggestions > would be greatly appreciated. Below is my input file: For systems with a non-standard electronic structure it might take 143 steps to converge. If it converges it will not be inaccurate. Systems like this are best treated using different approaches, like starting out with PCG and switching later to ODIIS. Usually we only need on wavefunction optimization from an initial guess. Later on (in geometry optimization or BOMD) you can start from much better guesses from previous calculations. They usually converge much quicker. BTW you are using Martins-Trouiller pseudos, not Vanderbilt. regards Juerg Hutter > > > Matt Wagner > Cleveland State University > Chemistry Dept. > mattwagner at hotmail.com > > &CPMD > WAVEFUNCTION OPTIMIZATION > PCG MINIMIZE > TIMESTEP > 20 > &END > > &DFT > FUNCTIONAL PBE > &END > > &SYSTEM > SYMMETRY > 0 > CHARGE > 1.0 > CELL > 40.0000000 1.0000000 1.0000000 0.0 0.0 0.0 > CUTOFF > 80 > &END > > &ATOMS > *H_MT_PBE.psp KLEINMAN-BYLANDER > LMAX=S > 15 > 10.2206634 20.0283459 17.1559610 > 10.1261771 16.8725020 17.1559610 > 14.6615216 19.6881951 13.7922472 > 21.9369701 21.3700520 18.4409753 > 22.6928608 16.6268375 21.2566684 > 20.3118049 21.1999766 23.9211833 > 20.8220312 18.1008245 24.9794303 > 16.7780157 17.9118518 24.0912587 > 25.9053965 19.3480442 19.8960650 > 26.1510610 23.3731625 21.3511547 > 24.6959713 19.6692978 25.4707593 > 27.9651988 16.6646321 25.4518620 > 29.4958776 20.3307022 26.2077528 > 29.8738229 20.1228322 23.0141144 > 28.0029933 22.7684499 24.5636904 > *O_MT_PBE.psp KLEINMAN-BYLANDER > LMAX=P > 3 > 18.9889961 20.7653394 14.0568090 > 24.4881013 22.5983745 21.3133602 > 26.7179791 16.7024266 24.1101560 > *C_MT_PBE.psp KLEINMAN-BYLANDER > LMAX=P > 9 > 13.7166581 18.3464890 17.3071392 > 17.8929545 19.7826814 15.8709468 > 18.8189207 18.9512016 18.2331054 > 17.0614747 18.5165644 20.1039350 > 22.2393264 18.6299480 21.2377711 > 20.3307022 19.2157633 23.3353679 > 24.7904576 19.9527568 21.5212301 > 26.0943692 19.2913524 24.0156696 > 28.5321169 20.7842367 24.4314095 > *N_MT_PBE.psp KLEINMAN-BYLANDER > LMAX=P > 5 > 14.5292407 17.9874409 19.6315033 > 11.1277323 18.4220781 16.8158102 > 15.3229260 19.2157633 15.5307959 > 21.3322575 19.5748115 18.8000234 > 17.8929545 18.2142081 22.5416827 > &END > > > _______________________________________________ > CPMD-list mailing list > CPMD-list at cpmd.org > http://cpmd.org/mailman/listinfo/cpmd-list > From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Wed Aug 24 17:50:26 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Wed, 24 Aug 2005 17:50:26 +0200 (CEST) Subject: [CPMD-list] 143 steps too long? In-Reply-To: Message-ID: On Tue, 23 Aug 2005, Matt wagner wrote: matt, MW> I am doing single point calculation for density minimization on BH4, MW> cation. Everything worked fined, but took 143 steps to complete. Is this MW> too long? Incorrect? I first used ODIIS and it took more than 1000 steps too long really depends on the system, that you are looking at. MW> for wavefunction optimization. When I used PCG Minimizer, that is when I MW> got it down to 143 steps. This was all done using the Vanderbelt MW> psuedopotential. The manual says that steps larger than 100 is to be MW> considered slow. Does slow mean inaccurate in this case? Any suggestions in some cases this is really difficult to tell. you can have a seemingly converged wavefunction, do an MD for quite a while and then suddenly your system drops into a different (lower) state. but this is a really rare event. in my personal experience, most of the time, however, slow convergence is due to one or some of the following reasons: - you have a 'problematic' element (e.g. many d-metals) -> here you have not much of a choice, sometimes switching to a different type of pseudopotential or switching between different methods to optimize the wavefunction (PCG, ODIIS, ANNEALING, LANCZOS) helps. - your converged wavefunction is very different from the initial (atomic) guess. -> you can try either using INITIALIZE WAVEFUNCTION RANDOM, or modify the initial guess using an ATOMIC CHARGES block in the &ATOMS section. - your starting structure is very 'bad'. -> if you just want to start an MD from the wavefunction, you can limit the number of wavefunction optimization steps. then use the not too well optimized wavefunction for a few steps of MD. then restart from that wavefunction and reoptimize to a tighter convergence. the same works, of course, for a geometry optimization. if you need the single point energy for _exactly_ your starting configuration, then you still could do this procedure and then, provided your atoms did not move too much, in the final step restart only from the wavefunction and _not_ from the coordinates, so that you can use the 'improved' wavefunction as initial guess. - you pick up a lot of numerical noise in the gradient correction. this is usually worst when using a low plane wave cutoff and for systems with a lot of vacuum. -> in this case you can change the GC-CUTOFF value from the default of 1.0e-8 to 1.0e-6 or what ever is needed. not that if you raise it too much, you will basically turn off the gradient correction, so you want to keep this number rather low. -> also in case of very low plane wave cutoffs, you need to increase the density cutoff. this can be done either directly via the CUTOFF DENSITY keyword or indirectly via the DUAL keyword (which defines the ratio between them). the defaul is a ratio of 4, but for a plane wave cutoff of 25 you should try raising it to 5 or 6. -> finally, you can reduce the noise by adapting the shape of the box and realigning the atoms in a way that there as little vaccuum area as possible, within the requirements of the poisson solver. - something else is wrong in your input. MW> would be greatly appreciated. Below is my input file: a few more remarks about your input file. MW> &SYSTEM MW> SYMMETRY MW> 0 MW> CHARGE MW> 1.0 hmmm. with a charge +1 you have an odd number of electrons in your system. if that is intentional, you have to use LSD. this may contribute to your convergence problems. MW> CELL MW> 40.0000000 1.0000000 1.0000000 0.0 0.0 0.0 from the coordinates it seems, that your molecule is much longer in x-direction than in the other directions, so using an orthorhombic box, might help cutting down the unneeded vacuum area, e.g. with CELL 40.0 0.8 0.8 0.0 0.0 0.0 and you can probably shrink the whole box a little, too, without losing accuracy (e.g. go from a=40 to a=38). MW> CUTOFF MW> 80 MW> &END since you don't have any problematic elements in your system i would expect an input with those two suggested modifications in place to converge within 50 steps or less. if you switch back to ultrasoft pseudopotentials with the associated additional changes, it may take a few steps more, but there should be no problems either. best 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 leiyinghui1977 at yahoo.com.cn Thu Aug 25 08:02:04 2005 From: leiyinghui1977 at yahoo.com.cn (yinghui lei) Date: Thu, 25 Aug 2005 14:02:04 +0800 (CST) Subject: [CPMD-list] Compiling using IFC and MKL Message-ID: <20050825060204.20780.qmail@web15608.mail.cnb.yahoo.com> Dear all , i have tried CPMD code on pentium4 using IFC compiler(8.2) and MKL library. i have some errors ...... reference to `_MKL_SERV_lsame' /net/mkl/users/sv/721/052/lapack/kernel/source/zunm2l.f:(.text+0x6f): undefined reference to `_MKL_SERV_lsame' /net/mkl/users/sv/721/052/lapack/kernel/source/zunm2l.f:(.text+0xab): undefined reference to `_MKL_SERV_lsame' /opt/intel/mkl721/lib/32/libmkl_lapack.a(zunm2r.o): In function `_MKL_LAPACK_zunm2r': /net/mkl/users/sv/721/052/lapack/kernel/source/zunm2r.f:(.text+0x3c): undefined reference to `_MKL_SERV_lsame' /opt/intel/mkl721/lib/32/libmkl_lapack.a(zunm2r.o):/net/mkl/users/sv/721/052/lapack/kernel/source/zunm2r.f:(.text+0x53): more undefined references to `_MKL_SERV_lsame' follow /usr/lib/libpthread.a(ptw-write.o): In function `write': :(.text+0x25): undefined reference to `__syscall_error' :(.text+0x55): undefined reference to `__syscall_error' /usr/lib/libpthread.a(ptw-read.o): In function `read': :(.text+0x25): undefined reference to `__syscall_error' :(.text+0x55): undefined reference to `__syscall_error' /usr/lib/libpthread.a(ptw-close.o): In function `close': :(.text+0x1f): undefined reference to `__syscall_error' /usr/lib/libpthread.a(ptw-close.o)::(.text+0x4a): more undefined references to `__syscall_error' follow make: *** [cpmd.x] ???? 1 My Makefile is: #---------------------------------------------------------------------------- # Makefile for cpmd.x (plane wave electronic calculation) # Configuration: PC-IFC # Creation of Makefile: 8?? 23 2005 # on Linux tiger 2.6.8-1-386 #1 Thu Nov 11 12:18:43 EST 2004 i686 GNU/Linux # Author: #---------------------------------------------------------------------------- # SHELL = /bin/sh # #--------------- Default Configuration for PC-IFC --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -pc64 -unroll -w90 -w95 -c -r8 -tpp7 -cm -tune pn4 -arch pn4 LFLAGS = /opt/intel/mkl721/lib/32/libmkl_lapack.a \ /opt/intel/mkl721/lib/32/libatlas_x86_64.a \ /opt/intel/mkl721/lib/32/libatlas_p4.a \ -i-static -static-libcxa -Bstatic -lsvml \ -Vaxlib -lguide -Xlinker -lpthread # $(QMMM_LIBS) FFLAGS_GROMOS= IRAT=2 CFLAGS = -c -O2 -Wall CPP = /lib/cpp -P -C -traditional CPPFLAGS = -D__Linux -D__PGI -DLAPACK -DFFT_DEFAULT -DLINUX_IFC \ -DADD_BLAS_ONE_UNDERSCORE CC = gcc -O2 -Wall FC = ifc -c -O3 -tp p6 -aixM LD = ifc -O3 -tp p6 -axiM AR = #---------------------------------------------------------------------------- CFGDEST = . CFGMACH = PC-IFC # # LIST OF FILES # OBJECTS = $(OBJ_AL) $(OBJ_WF) $(OBJ_GM) $(OBJ_MD) $(OBJ_KS) \ ...... may be anybody know where is wrong any info is acknowledged best regards Yongping ZENG ___________________________________________________________ ????????G?????????????????????????????????????? http://cn.mail.yahoo.com From L.Walewski at icm.edu.pl Thu Aug 25 12:36:22 2005 From: L.Walewski at icm.edu.pl (=?ISO-8859-2?Q?=A3ukasz_Walewski?=) Date: Thu, 25 Aug 2005 12:36:22 +0200 Subject: [CPMD-list] PIMD: EHAM depends on number of processors ? Message-ID: <430D9F26.2040805@icm.edu.pl> Dear CPMD users, When playing with path integral MD I observed strange behaviour: the EHAM value is conserved during the simulation (as it should be) however its mean value depends on the number of processors the job runs on. I run several tests with the following results: version, arch, no. of proc., EHAM (a.u.) 3.9.1, Intel, NP=1: -49.9124 3.9.2, Intel, NP=1: -49.9124 3.9.2, Cray, NP=1: -49.9130 3.9.1, Cray, NP=4: -49.9080 3.9.2, Cray, NP=4: -49.9080 3.9.1, Opteron, NP=1: -49.9117 3.9.1, Opteron, NP=4: -49.9066 3.9.1, Opteron, NP=8: -49.9166 The largest difference occurs for 3.9.1 on the Opteron cluster between 4 and 8 processors: 6.3 kcal/mol ! Version number does not not play any role (at least on serial PC and Cray). Is it due to numerical errors (like order of summation and other known nondeterministic aspects of parallel computing) or do you suspect problems somewhere else ? Are the results feasible ? correct ? Are other properties of the system affected ? It may be important that all above simulations were carried out from the same input file attached below (most of it was inspired by CPMD-test/path-int/inp-2). The wavefunction was initially optimised and the configuration of the centroids was generated using DEBROGLIE CENTROID scheme. With warm regards, Lukasz &CPMD PATH INTEGRAL MOLECULAR DYNAMICS CP ODIIS 5 RESTART WAVEFUNCTION COORDINATES VELOCITIES LATEST ISOLATED MOLECULE EMASS 350.0 TEMPERATURE 300.D0 NOSE IONS MASSIVE 300.0 2500.0 NOSE ELECTRONS 2.D-4 15000.D0 TIMESTEP 1.D0 MAXSTEP 13781 TRAJECTORY XYZ SAMPLE 13 &END &SYSTEM ANGSTROM SYMMETRY 1 CELL 10.0 1.0 1.0 0 0 0 CUTOFF 35.0 &END &ATOMS *pp/O_MT_PBE.psp KLEINMAN-BYLANDER LMAX=P 2 1.184893 -0.892757 0.000000 -1.184893 -0.892757 0.000000 *pp/C_MT_PBE.psp KLEINMAN-BYLANDER LMAX=P 3 1.187504 0.388446 0.000000 0.000000 1.130651 0.000000 -1.187504 0.388446 0.000000 *pp/H_MT_PBE.psp KLEINMAN-BYLANDER LMAX=S 4 0.000000 -1.136055 0.000000 2.165818 0.881532 0.000000 0.000000 2.211832 0.000000 -2.165818 0.881532 0.000000 &END &DFT NEWCODE FUNCTIONAL PBE &END &PIMD TROTTER DIMENSION 8 NORMAL MODES 1.D0 FACMASS 1.D0 DEBROGLIE CENTROID 300.D0 PRINT LEVEL 1 &END -- -------------------------------- ?ukasz Walewski From g0403127 at nus.edu.sg Thu Aug 25 10:19:13 2005 From: g0403127 at nus.edu.sg (Hong Won Keon) Date: Thu, 25 Aug 2005 16:19:13 +0800 Subject: [CPMD-list] Pt TM_PBE pseudo-potential Message-ID: <9E4557B19235FE47B1C7B52DE3A536A33E04DB@MBOX23.stu.nus.edu.sg> Dear CPMD-lists, I produced Pt pseudo-potential using fhi98PP program. With this PP, I could get lattice constant of around 4.1Angstrom. But I met problem in atomic energy calculation. The thing is when the job is run, CPMD recognized only S angular quantum number, ignoring other P or D orbitals. I don't know what is wrong I'll appreciate anyone who can give some advice on this matter. INPUT files ************************* &INFO To get equilibrium lattice constant of Bulk Platinum single point calculation &END &CPMD OPTIMIZE WAVEFUNCTION CONVERGENCE ORBITALS 1.0e-7 LSD FREE ENERGY FUNCTIONAL TROTTER FACTOR 0.001 BOGOLIUBOV CORRECTION ON PRINT FORCES ON MAXSTEP 1000 &END &DFT FUNCTIONAL PBE &END &SYSTEM ANGSTROM SYMMETRY 1 SCALE CELL 12.0 1.0 1.0 0 0 0 CUTOFF 42.0 MULTIPLICITY 3 STATES 18 &END &ATOMS *Pt_PBE.psp KLEINMAN-BYLANDER LMAX = D LOC = P 1 0.0 0.0 0.0 &END &BASIS PSEUDO AO 2 0 2 &END part of OUTPUT ************************************************************ fhi pseudopotential tool fhipp - version rev270603B chemical symbol Pt nuclear charge 78.00 total charge 0.00 number of core states 12 number of valence states 2 exchange-correlation model 6 GGA Perdew/Burke/Ernzerhof scalar-relativistic mode parameters radial mesh 567 1.024700 0.801282E-04 l type rcore rmatch eigenvalue(eV) norm te all-electron pseudo 1 = 0 t 2.4934552 2.5550435 -5.6593889 -5.6594197 0.99999 1 t 2.6181531 2.6828215 -5.6593889 -5.6595313 0.99999 2 t 2.4934552 2.5550435 -6.1413403 -6.1413661 0.99999 ************************************************************ **************************************************************** * ATOM MASS RAGGIO NLCC PSEUDOPOTENTIAL * * Pt 195.0900 1.2000 NO S LOCAL * **************************************************************** PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA NCPU NGW NHG PLANES GXRAYS HXRAYS ORBITALS Z-PLANES 0 6712 53571 24 429 1717 4 1 1 6716 53591 24 432 1718 5 1 2 6716 53595 24 432 1718 4 1 3 6714 53588 24 432 1716 5 1 G=0 COMPONENT ON PROCESSOR : 0 PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA Thanks, HONG W.K. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050825/0dba8286/attachment.html From osirixu-cpmd at yahoo.com.cn Sun Aug 28 18:52:06 2005 From: osirixu-cpmd at yahoo.com.cn (osirixu-cpmd at yahoo.com.cn) Date: Mon, 29 Aug 2005 00:52:06 +0800 (CST) Subject: [CPMD-list] Is there any problem in this Makefile? In-Reply-To: <200405120955.LAA27180@cpmd.org> Message-ID: <20050828165206.48562.qmail@web15802.mail.cnb.yahoo.com> I have installed the latest Intel Fortran Complier (v9.0.021), Intel C++ Complier (v9.0.021), and Intel Math Kernel Library (v7.2.1.003) on my CentOS4 (a clone package of RHEL4). My computer processor is athlon K5 1050Mhz. I wonder if there is any problem in the CPMD-3.9.2 Makefile below for my computer, or is there any tips to optimize this configuration? # SHELL = /bin/sh # #--------------- Default Configuration for PC-IFC --------------- SRC = . DEST = . BIN = . #QMMM_FLAGS = -D__QMECHCOUPL #QMMM_LIBS = -L. -lmm FFLAGS = -c -r8 -O3 -pc64 -xW -tpp6 -unroll -cm -tune pn3 -w90 -w95 LFLAGS = -lsvml -L/opt/intel/mkl721/lib/32/ -lmkl_lapack -lmkl_ia32 -Vaxlib /opt/intel/mkl721/lib/32/libguide.a -lpthread $(QMMM_LIBS) CFLAGS = -c -O2 -Wall CPP = /lib/cpp -P -C -traditional CPPFLAGS = -D__Linux -D__PGI -DFFT_DEFAULT -DLINUX_IFC -DINTEL_MKL NOOPT_FLAG = CC = icc -c -O FC = ifc LD = ifc AR = ar #---------------------------------------------------------------------------- Besides, I have changed the util.F file around line 550 from: #if !(defined(__SGI)) && !(defined(__HP)) && !(defined(LINUX_IA64_INTEL)) to: #if !(defined(__SGI)) && !(defined(__HP)) && !(defined(LINUX_IFC)) p.s. Would it be better if using -lmkl_lapack rather than -lmkl_lapack32, and -lmkl_ia32 rather than -lmkl_p3 in LFLAGS here? Indeed, I am a little confused about the settings for FFLAGS and LFLAGS. Is there any detailed parameter setting description of the two flags, respectively for specific computer processor and compile environment? Many thanks! --------------------------------- DO YOU YAHOO!? ????????G?????????????????????????????????????? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050829/a24d87a9/attachment.html From ari.p.seitsonen at iki.fi Sun Aug 28 20:02:24 2005 From: ari.p.seitsonen at iki.fi (Ari P Seitsonen) Date: Sun, 28 Aug 2005 20:02:24 +0200 (CEST) Subject: [CPMD-list] Pt TM_PBE pseudo-potential In-Reply-To: <9E4557B19235FE47B1C7B52DE3A536A33E04DB@MBOX23.stu.nus.edu.sg> References: <9E4557B19235FE47B1C7B52DE3A536A33E04DB@MBOX23.stu.nus.edu.sg> Message-ID: Dear Hong W.K., Please don't leave a space after the '=' sign after 'LMAX'; it's allowed to have LMAX=D LMAX =D LMAX =D but not e.g. LMAX= D The same applies for 'LOC=...'. Greetings, apsi -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=- Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/ CNRS & IMPMC, Universit? Pierre et Marie Curie 4 place Jussieu, case 115 / F-75252 Paris Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820 On Thu, 25 Aug 2005, Hong Won Keon wrote: > Dear CPMD-lists, > > > > I produced Pt pseudo-potential using fhi98PP program. > > With this PP, I could get lattice constant of around 4.1Angstrom. > > But I met problem in atomic energy calculation. > > The thing is when the job is run, CPMD recognized only S angular quantum > number, ignoring other P or D orbitals. > > I don't know what is wrong > > I'll appreciate anyone who can give some advice on this matter. > > > > > > INPUT files > > ************************* > > &INFO > To get equilibrium lattice constant of Bulk Platinum > single point calculation > &END > > > &CPMD > OPTIMIZE WAVEFUNCTION > CONVERGENCE ORBITALS > 1.0e-7 > LSD > FREE ENERGY FUNCTIONAL > TROTTER FACTOR > 0.001 > BOGOLIUBOV CORRECTION ON > PRINT FORCES ON > MAXSTEP > 1000 > &END > > > &DFT > FUNCTIONAL PBE > &END > > > &SYSTEM > ANGSTROM > SYMMETRY > 1 > SCALE > CELL > 12.0 1.0 1.0 0 0 0 > CUTOFF > 42.0 > MULTIPLICITY > 3 > STATES > 18 > &END > > > &ATOMS > *Pt_PBE.psp KLEINMAN-BYLANDER > LMAX = D LOC = P > 1 > 0.0 0.0 0.0 > &END > > > &BASIS > PSEUDO AO 2 > 0 2 > &END > > > > > part of OUTPUT > > ************************************************************ > fhi pseudopotential tool fhipp - version rev270603B > > chemical symbol Pt > nuclear charge 78.00 > total charge 0.00 > number of core states 12 > number of valence states 2 > exchange-correlation model 6 GGA Perdew/Burke/Ernzerhof > scalar-relativistic mode > parameters radial mesh 567 1.024700 0.801282E-04 > > l type rcore rmatch eigenvalue(eV) norm te > all-electron pseudo 1 = > 0 t 2.4934552 2.5550435 -5.6593889 -5.6594197 0.99999 > 1 t 2.6181531 2.6828215 -5.6593889 -5.6595313 0.99999 > 2 t 2.4934552 2.5550435 -6.1413403 -6.1413661 0.99999 > ************************************************************ > > **************************************************************** > * ATOM MASS RAGGIO NLCC PSEUDOPOTENTIAL * > * Pt 195.0900 1.2000 NO S LOCAL * > **************************************************************** > > > PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA > NCPU NGW NHG PLANES GXRAYS HXRAYS ORBITALS Z-PLANES > 0 6712 53571 24 429 1717 4 1 > 1 6716 53591 24 432 1718 5 1 > 2 6716 53595 24 432 1718 4 1 > 3 6714 53588 24 432 1716 5 1 > G=0 COMPONENT ON PROCESSOR : 0 > PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA > > > > > Thanks, > > HONG W.K. > > > > From axel.kohlmeyer at theochem.ruhr-uni-bochum.de Sun Aug 28 22:38:48 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Sun, 28 Aug 2005 22:38:48 +0200 (CEST) Subject: [CPMD-list] Compiling using IFC and MKL In-Reply-To: <20050825060204.20780.qmail@web15608.mail.cnb.yahoo.com> Message-ID: On Thu, 25 Aug 2005, yinghui lei wrote: YL> Dear all , i have tried CPMD code on pentium4 using YL> IFC compiler(8.2) and MKL library. i have some errors please note, that the last intel fortran releases were version 7.1, 8.0, 8.1, and 9.0. so you either must have a different version or a fake. ;-) YL> reference to `_MKL_SERV_lsame' YL> /net/mkl/users/sv/721/052/lapack/kernel/source/zunm2l.f:(.text+0x6f): YL> undefined reference to `_MKL_SERV_lsame' well you are obviously linking the LAPACK part of MKL but not the blas/core part of MKL. [...] YL> :(.text+0x1f): undefined reference to YL> `__syscall_error' YL> /usr/lib/libpthread.a(ptw-close.o)::(.text+0x4a): YL> more undefined references to `__syscall_error' follow YL> make: *** [cpmd.x] ???? 1 seems like you have an inconsistent versions of the c-library installed as well. in case you have intel 7.1 you'll probably need to update to a newer version. YL> FFLAGS = -pc64 -unroll -w90 -w95 -c -r8 -tpp7 -cm YL> -tune pn4 -arch pn4 YL> LFLAGS = /opt/intel/mkl721/lib/32/libmkl_lapack.a \ YL> /opt/intel/mkl721/lib/32/libatlas_x86_64.a \ YL> /opt/intel/mkl721/lib/32/libatlas_p4.a \ YL> -i-static -static-libcxa -Bstatic -lsvml \ YL> -Vaxlib -lguide -Xlinker -lpthread o dear! (sigh) please clean up your installation of MKL. you cannot mix-and-match totally different libraries like MKL and atlas and expect that to work. the two atlas binaries have no place in the mkl directory and furthermore the x86_64 version is totally misplaced in the 32-bit directory. also you are using a cruel patchwork of flags that are needed/supported for different compilers (-i-static is only supported with intel fortran 9.0 or recent updates to 8.1 from premier support, -static-libcxa is for older compilers, -Bstatic is a PGI compiler flag, also -lsvml and -Xlinker make little sense there. YL> # $(QMMM_LIBS) YL> FFLAGS_GROMOS= YL> IRAT=2 YL> CFLAGS = -c -O2 -Wall YL> CPP = /lib/cpp -P -C -traditional YL> CPPFLAGS = -D__Linux -D__PGI -DLAPACK -DFFT_DEFAULT YL> -DLINUX_IFC \ YL> -DADD_BLAS_ONE_UNDERSCORE this define does not exist in CPMD and thus totally pointless. YL> YL> CC = gcc -O2 -Wall YL> FC = ifc -c -O3 -tp p6 -aixM YL> LD = ifc -O3 -tp p6 -axiM these flags again inconsistent with the compiler flags from above where you want to optimize for a pentium-iv architecture. it would be helpful, if you'd state what kind of machine you are using and what version of linux exactly you have installed. please study the documentation of both mkl and the intel compiler carefully on how to use them properly and then modify your makefile accordingly. just putting compilerflags somewhere randomly only increases the chance for a miscompile. YL> AR = YL> #---------------------------------------------------------------------------- YL> CFGDEST = . YL> CFGMACH = PC-IFC YL> # YL> # LIST OF FILES YL> # YL> OBJECTS = $(OBJ_AL) $(OBJ_WF) $(OBJ_GM) $(OBJ_MD) YL> $(OBJ_KS) \ YL> ...... YL> YL> may be anybody know where is wrong YL> any info is acknowledged YL> YL> best regards YL> Yongping ZENG one more suggestion: it would be less confusing, if you'd use an email account that carries the same name, that you are signing your mails with.... best regards, axel kohlmeyer. YL> YL> ___________________________________________________________ YL> ????????G?????????????????????????????????????? YL> http://cn.mail.yahoo.com YL> YL> _______________________________________________ YL> CPMD-list mailing list YL> CPMD-list at cpmd.org YL> http://cpmd.org/mailman/listinfo/cpmd-list YL> YL> -- ======================================================================= 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 Sun Aug 28 22:46:30 2005 From: axel.kohlmeyer at theochem.ruhr-uni-bochum.de (Axel Kohlmeyer) Date: Sun, 28 Aug 2005 22:46:30 +0200 (CEST) Subject: [CPMD-list] Pt TM_PBE pseudo-potential In-Reply-To: <9E4557B19235FE47B1C7B52DE3A536A33E04DB@MBOX23.stu.nus.edu.sg> Message-ID: On Thu, 25 Aug 2005, Hong Won Keon wrote: HWK> Dear CPMD-lists, dear hong, HWK> I produced Pt pseudo-potential using fhi98PP program. HWK> HWK> With this PP, I could get lattice constant of around 4.1Angstrom. HWK> HWK> But I met problem in atomic energy calculation. HWK> HWK> The thing is when the job is run, CPMD recognized only S angular quantum HWK> number, ignoring other P or D orbitals. sorry if i am missing something obvious here, but would you mind explaining, how you get the lattice constant, but fail to run a single point calculation? HWK> HWK> I don't know what is wrong HWK> HWK> I'll appreciate anyone who can give some advice on this matter. [...] HWK> &SYSTEM HWK> ANGSTROM HWK> SYMMETRY HWK> 1 HWK> SCALE HWK> CELL HWK> 12.0 1.0 1.0 0 0 0 HWK> CUTOFF HWK> 42.0 HWK> MULTIPLICITY HWK> 3 HWK> STATES HWK> 18 HWK> &END another stupid question: you find a lattice contant of 4.1 angstrom, yet you use 12.a.u. and a _primitive_ cubic cell. i'd have expected 7.8a.u. and FCC symmetry (IBRAV=2) since you give only one atom. for IBRAV=1 you'd have to give _all_ coordinates for the conventional cell not only the primitive cell. also, i'd expect that you'd use k-points. HWK> HWK> &ATOMS HWK> *Pt_PBE.psp KLEINMAN-BYLANDER HWK> LMAX = D LOC = P here is your main problem. the syntax is: LMAX=D LOC=P there must not be any blanks! [...] HWK> chemical symbol Pt HWK> nuclear charge 78.00 HWK> total charge 0.00 HWK> number of core states 12 HWK> number of valence states 2 HWK> exchange-correlation model 6 GGA Perdew/Burke/Ernzerhof HWK> scalar-relativistic mode HWK> parameters radial mesh 567 1.024700 0.801282E-04 HWK> HWK> l type rcore rmatch eigenvalue(eV) norm te HWK> all-electron pseudo 1 = HWK> 0 t 2.4934552 2.5550435 -5.6593889 -5.6594197 0.99999 HWK> 1 t 2.6181531 2.6828215 -5.6593889 -5.6595313 0.99999 HWK> 2 t 2.4934552 2.5550435 -6.1413403 -6.1413661 0.99999 HWK> ************************************************************ HWK> **************************************************************** HWK> * ATOM MASS RAGGIO NLCC PSEUDOPOTENTIAL * HWK> * Pt 195.0900 1.2000 NO S LOCAL * HWK> **************************************************************** i have yet one more comment: you are using a minimal pseudopotential without NLCC for a heavy element. this may even work to some degree (i don't know how much accuracy you want) for the bulk metal, but most likely you'll need a semi-core pseudopotential. best regards, axel kohlmeyer. HWK> HWK> HWK> PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA HWK> NCPU NGW NHG PLANES GXRAYS HXRAYS ORBITALS Z-PLANES HWK> 0 6712 53571 24 429 1717 4 1 HWK> 1 6716 53591 24 432 1718 5 1 HWK> 2 6716 53595 24 432 1718 4 1 HWK> 3 6714 53588 24 432 1716 5 1 HWK> G=0 COMPONENT ON PROCESSOR : 0 HWK> PARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARAPARA HWK> HWK> HWK> HWK> HWK> Thanks, HWK> HWK> HONG W.K. HWK> HWK> 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 g0403127 at nus.edu.sg Mon Aug 29 09:49:49 2005 From: g0403127 at nus.edu.sg (Hong Won Keon) Date: Mon, 29 Aug 2005 15:49:49 +0800 Subject: [CPMD-list] Carbon and Hydrogen LMAX & LOC setting Message-ID: <9E4557B19235FE47B1C7B52DE3A536A33E04DE@MBOX23.stu.nus.edu.sg> Dear CPMD-lists, I'd like to calculate hydro-carbon molecule, so I need Carbon and Hydrogen Pseudo Potentials. As we know well, in the PP library of CPMD there are many kinds of PPs. I chose SG & MT PP with BLYP & PBE functionals. But what I am confused it LMAX and LOC of Carbon. For MT PP, I'd like to use KLEINMAN-BYLANDER method for non-local part integration. So I need to know which angular momentum should be set as local. At MT PP files, they implements by D angular momentum, but Carbon has 1S2 2S2 2P2 system. So should I set LMAX to D? Or LMAX=P? When I set LMAX=P means, P is considered as local, on the other side, if I set S as local, in both case KB method might lose one angular momentum. For SG PP, I heard that no need of above non-local part integration method. Then is it mean the LMAX should be set as D. Also, for C_MT_BLYP PP case, there are two - C_MT_BLYP_1.psp & C_MT_BLYP_2.psp Which one is better? Or any difference? The problem I met is when I calculated C atomic energy with C_MT_BLYP_2.psp setting LMAX=P, the total energy was -10.xxxx But with C_MT_PBE.psp with LMAX=P LOC=S, I could get the total energy of -5.xxxx. The both SG PPs give the total energy of -5.xxxx. I think the problem is on the LMAX and LOC setting. Another stupid question is about Hydrogen. I think it should be set LMAX=S in any kinds of PP. But still MT PP shows another momentum of P. to be sure, I ask this. In summary, Carbon and Hydrogen PP, which angular momentum should be set LMAX and LOC? The PPs I am using is C_MT_BLYP_2.psp, C_MT_PBE.psp, C_SG_BLYP & C_SG_PBE H_MT_BLYP.psp, H_MT_PBE.psp, C_SG_BLYP & C_SG_PBE Thanks in advance, HONG W.K. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cpmd.org/pipermail/cpmd-list/attachments/20050829/5cf53c0f/attachment.html From mclar at elrond.chem.uni.wroc.pl Tue Aug 30 12:40:45 2005 From: mclar at elrond.chem.uni.wroc.pl (Przemyslaw Dopieralski) Date: Tue, 30 Aug 2005 12:40:45 +0200 (CEST) Subject: [CPMD-list] TESR with water molecules calculation Message-ID: Hi :) I'm trying to calculate ion with water molecules should I use TESR option? Przemyslaw Dopieralski Ph.D. student University od Wroclaw POLAND e-mail: mclar at elrond.chem.uni.wroc.pl