[gmx-users] Parameter file - dt and nsteps
anoddlad at yahoo.com
Tue Oct 18 16:09:08 CEST 2005
Time in the parameter file is in ps... so dt=2 will be
2000 fs, and 0.002 will be the 2 fs that you want. 3
ns will naturally take a while. mdrun will usually
tell you when it will finish. It depends on the size
of your system and speed of your processor, of course,
but generally you're looking at days, possibly weeks
for an MD run in the order of several nanoseconds. I
have a system size around 21000 particles and get
about half a nanosecond a day over 8 processors, using
a 1.5fs timestep. You can tweak the speed by
modifying the cutoffs etc, just don't reduce them too
--- Fernando Mattio <mattiofer at gmail.com> wrote:
> Dear Gromacs users,
> I want to perform simulations of 3 ns with gromacs,
> and I believe that the
> simulation time is determined in the parameter file
> (dt and nsteps). I
> performed simulation of 0,1 ns with dt=0,002 and
> nsteps =50000, and it was
> ok after 100 minutes. So I am really worried about
> how much time a 3 ns
> simulation can takes!
> I realized that if I changed the dt it could go
> faster (with less time
> steps), but every time I did it (dt = 0,004, dt =
> 0,02, dt = 0,004, dt = 2)
> my system collapsed!
> It consists of a protein in a box of water... And I
> read somewhere in the
> gromacs forum that for a system (protein and
> solvent) like this I could use
> at most dt = 2 femtoseconds.
> So shouldn´t it work?
> The Fatal error was always:
> Number of grid cells is zero. Probably the system
> and box collapsed.
> Any information or advices about would be nice
> because I am new user of
> Or the only way is to be very very patient?
> Thank you very much in advance,
> Fernando Mattio
> > _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> Please don't post (un)subscribe requests to the
> list. Use the
> www interface or send it to
gmx-users-request at gromacs.org.
Start your day with Yahoo! - Make it your home page!
More information about the gromacs.org_gmx-users