Re: [gmx-users] run gromacs on a single machine with mutiple CPUs
Yang Ye wrote: normally we use following commands when compiling gromacs ./configure ... --program_suffix=_mpi --enable-mpi ... So we will have mdrun_mpi in the end. The usage of mdrun_mpi is the same as mdrun but -np, -replex and other switches are activated. That statement on the user manual is to clarify that you don't need MPI-enabled mdrun to run on a multiprocessor machine. But if you want to tap the capability of the multiprocessor machine, you need to use MPI and use mdrun_mpi As Erik stated, different MPI implementations have different standards for starting jobs. The comment in the manual is probably a remnant from an old SGI MPI implementation. I will remove the comment. Otherwise, a little trial and error and man XXX usually does the job... -- David. David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group, Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596, 75124 Uppsala, Sweden phone: 46 18 471 4205 fax: 46 18 511 755 [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] run gromacs on a single machine with mutiple CPUs
Thanks for your help very much. Best, Ocean On 12/21/06, Yang Ye [EMAIL PROTECTED] wrote: normally we use following commands when compiling gromacs ./configure ... --program_suffix=_mpi --enable-mpi ... So we will have mdrun_mpi in the end. The usage of mdrun_mpi is the same as mdrun but -np, -replex and other switches are activated. That statement on the user manual is to clarify that you don't need MPI-enabled mdrun to run on a multiprocessor machine. But if you want to tap the capability of the multiprocessor machine, you need to use MPI and use mdrun_mpi Regards, Yang Ye On 12/21/2006 4:20 PM, Seaclear Theory wrote: On 12/20/06, *Mark Abraham* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Seaclear Theory wrote: On 12/20/06, *Mark Abraham* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Seaclear Theory wrote: Hi! All, I have a linux server with 4 CPUs. How can I run gromacs on all CPUs? The gromacs manual has very limit information. And it does not works. So which part of the manual section entitled Running GROMACS in parallel doesn't work? What I want is to run gromacs on a signle machine with multiple processors. In A.5 Running GROMACS in parallel If you have a single machine with multiple processors you don't have to use the mpirun command, but you can do with an extra option to mdrun: % mdrun -np 8 -s topol -v -N 8 In this example MPI reads the first option from the command line. Since mdrun also wants to know the number of processes you have to type it twice. So what does it say? Should you be using mdrun_mpi? How about using a mpirun command anyway... If you want meaningful feedback you need to describe your setup thoroughly, say what you've tried and why that didn't work and why the error messages didn't help you. Otherwise, you need to pay someone for a support contract :-) I'll help solve problems, but I won't help if you don't give the impression you're helping yourself too! Mark I do not understand why you are so upset. Are you the author of the user manual :-) Any way, let's go back for the issue, not person. The user manual said that you don't have to use the mpirun command if you want to run gromacs on a single machine with multiple processors. How should I understand this? By the way, the manual ( v.3.3) and the reference does not have information about mdrun_mpi. Thanks. Ocean ___ gmx-users mailing listgmx-users@gromacs.org mailto:gmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] mailto:[EMAIL PROTECTED]. Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] run gromacs on a single machine with mutiple CPUs
Seaclear Theory wrote: I believed the case that single machine with multiple processors (like Dell server with 2 due core CPUs) should be very popular for end user. Because not everyone can offer a 32 nodes cluster or has access to supercomputer center. Could we have detail tutorial for run gromacs on single machine with multiple processors Thanks. Best, By now you know how to do it, so if you write the tutorial (also for clusters please) I will put it on the website. -- David. David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group, Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596, 75124 Uppsala, Sweden phone: 46 18 471 4205 fax: 46 18 511 755 [EMAIL PROTECTED] [EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] run gromacs on a single machine with mutiple CPUs
Will work on it. Thanks. Ocean On 12/21/06, David van der Spoel [EMAIL PROTECTED] wrote: Seaclear Theory wrote: I believed the case that single machine with multiple processors (like Dell server with 2 due core CPUs) should be very popular for end user. Because not everyone can offer a 32 nodes cluster or has access to supercomputer center. Could we have detail tutorial for run gromacs on single machine with multiple processors Thanks. Best, By now you know how to do it, so if you write the tutorial (also for clusters please) I will put it on the website. -- David. David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group, Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596, 75124 Uppsala, Sweden phone: 46 18 471 4205 fax: 46 18 511 755 [EMAIL PROTECTED][EMAIL PROTECTED] http://folding.bmc.uu.se ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
Re: [gmx-users] run gromacs on a single machine with mutiple CPUs
Seaclear Theory wrote: I do not understand why you are so upset. Are you the author of the user manual :-) Any way, let's go back for the issue, not person. I'm not upset - I'm just communicating directly :-) I do get a little irritated when someone wanting help doesn't describe their problem very well, but sometimes it is hard for them to know what to describe :-) Asserting the manual doesn't describe much and it's wrong anyway just makes most of us think you haven't looked hard enough and are trying to get us to do all of your work. Saying I read A.5 in the manual, and when I tried stuff, then weird thing this happened... what's up? is far better for everyone. The user manual said that you don't have to use the mpirun command if you want to run gromacs on a single machine with multiple processors. How should I understand this? By the way, the manual ( v.3.3) and the reference does not have information about mdrun_mpi. Thanks. OK. configure --enable-mpi creates versions of all the executables that have an _mpi suffix... only mdrun actually needs this, however, so the usual best strategy is to do a configure without the MPI, install that, then configure with enable-mpi and then make install-mdrun. When you want a multi-processor mdrun, use the appropriate one.. If mdrun_mpi doesn't work as in A.5, then I suggest using an mpirun version anyway, since with the correct MPI configuration (and you're on your own here) you can run any combination of processes over processors. Cheers, Mark ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] do_dssp
Dear Gromacs users, I have simulated 12 peptide fragments in a cubic box and I want to use do_dssp for secondary structure analysis but I am not sure whether Periodic boundary conditions will be taken in account during assignment (particularly for assigning betasheets). Regards, Gurpreet Singh - University of Dortmund Department of Chemistry Physical Chemistry I - Biophysical Chemistry Otto-Hahn Str. 6 D-44227 Dortmund Germany Office: C1-06 room 176 Phone: +49 231 755 3916 Fax: +49 231 755 3901 - ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php
[gmx-users] limit of number of processors ??
Hello All, I have been doing membranes simulations using the computer nodes in UTEXAS supercomputing center. Each board contains two Xeon Intel Duo-Core 64-bit processors (4 cores in all) and nodes are interconnected with InfiniBand technology. The performance is great. However, I have got weird results if I increase the number of processors more than 12. Even though the bilayer is still intact, one or two lipids were popped out of the membrane not long after the simulation was started. To make sure that it is not the problem due to different initial conditions, I ran two separate md simulations starting from the same structure and velocities on each atom but only with different number of processors, 8 and 16. As expected, the job with 8 processors resulted in the intact membrane but a lipid was pulled out of the membrane in the job run with 16 processors. So far, I could not find out the cause of this problem. Does anyone have any thought about this problem? Thank you Sunjoo _ 메신저에서 문자를 바로 보내보세요 http://phonebuddy.msn.co.kr/ ___ gmx-users mailing listgmx-users@gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post (un)subscribe requests to the list. Use the www interface or send it to [EMAIL PROTECTED] Can't post? Read http://www.gromacs.org/mailing_lists/users.php