[gmx-users] nsgrid error

[Mark Abraham]

Attilio Vargiu wrote:
> Hi All,
> I'm simulating a drug-DNA complex in water, with the AMBER force field and
> running GROMACS in parallel on 8 processors.
> After ~100 ps, I get the error here below:
> 
> ---------
> Program mdrun, VERSION 3.3.1
> Source code file: nsgrid.c, line: 226
> 
> Range checking error:
> Explanation: During neighborsearching, we assign each particle to a grid
> based on its coordinates. If your system contains collisions or parameter
> errors that give particles very high velocities you might end up with some
> coordinates being +-Infinity or NaN (not-a-number). Obviously, we cannot
> put these on a grid, so this is usually where we detect those errors.
> Make sure your system is properly energy-minimized and that the potential
> energy seems reasonable before trying again.
> 
> Variable ci has value 3035. It should have been within [ 0 .. 2940 ]
> Please report this to the mailing list (gmx-users at gromacs.org)
> ---------
> 
> It is not the first time I got this error, but I'm pretty sure the system
> is well equilibrated (I taken the starting structure from a previous 3ns
> long simulation, I do not see anything strange visualizing the traj, and
> the solute RMSD with respect to the initial configuration stays below 1.5
> angs), and the potential energy is reasonable and with no drift.
> Sometimes I tried simply to restart the simulation from the previous
> "stable" point with coordinates, velocities and energy information
> (through tpbconv), and this had fixed the problem.
> Is not this a little bit strange? Could be a problem of the compiler?
> I'm using the fortran compiler intel9.0, the gnu c compiler, mkl-7.2 and
> mpich.

Yeah that does sound weird. My first suspect would be MPICH - try LAM.

Mark