Re: [gmx-users] Problem on i7

2010-02-27 Thread Roland Schulz
Hi, Ubuntu has also gromacs-mpich and gromacs-openmpi which contain mdrun_mpi to run in parallel. Roland On Sat, Feb 27, 2010 at 1:53 PM, Maurício Menegatti Rigo < mauriciomr1...@gmail.com> wrote: > Hi Roland, > > thanks for your answer. Actually, I installed the GROMACS from Synaptic > package

Re: [gmx-users] Problem on i7

2010-02-27 Thread Maurício Menegatti Rigo
Hi Roland, thanks for your answer. Actually, I installed the GROMACS from Synaptic package manager. I dont know if the manager compile with the "--enable-mpi" option. I'll try to reinstall with this option and see what happens. Thank you very much!! -- gmx-users mailing listgmx-users@gromacs

Re: [gmx-users] Problem on i7

2010-02-27 Thread Roland Schulz
Hi, this happens if the mdrun hasn't been compiled with "configure --enable-mpi". Roland On Sat, Feb 27, 2010 at 1:04 PM, Maurício Menegatti Rigo < mauriciomr1...@gmail.com> wrote: > Hi, > > I'm facing a problem where the GROMACS can't distribute the job that I send > between the 8 nodes of my

[gmx-users] Problem on i7

2010-02-27 Thread Maurício Menegatti Rigo
Hi, I'm facing a problem where the GROMACS can't distribute the job that I send between the 8 nodes of my computer (i7 920). Actually, when I send the job with the command line "mpirun -np 8 mdrun (...)" each CPU receive the information and perform the job by itself, without distribution. Someone

Re: [gmx-users] correlation function 'time value' ?

2010-02-27 Thread Mark Abraham
On 27/02/2010 7:31 PM, Ramachandran G wrote: Dear Dr. van der Spoel: Thanks for the clarification in your reply. One more thing I was just curious is that, for C(t) = < f(x).f(x+t) >, what is the default value for t, which is often called "correlation time", in gromacs for correlation function c

Re: [gmx-users] correlation function 'time value' ?

2010-02-27 Thread David van der Spoel
On 2010-02-27 09.31, Ramachandran G wrote: Dear Dr. van der Spoel: Thanks for the clarification in your reply. One more thing I was just curious is that, for C(t) = < f(x).f(x+t) >, what is the default value for t, which is often called "correlation time", in gromacs for correlation function cal

Re: [gmx-users] correlation function 'time value' ?

2010-02-27 Thread Ramachandran G
Dear Dr. van der Spoel: Thanks for the clarification in your reply. One more thing I was just curious is that, for C(t) = < f(x).f(x+t) >, what is the default value for t, which is often called "correlation time", in gromacs for correlation function calculation, especially in "g_hbond", and/or is