[gmx-users] Help: Potential energy = nan

[Arturas]

Hello, dr. gmx-users,

There are problem porting my prepared MD files on cluster.

I simulate organic molecule clutser (MPI) in water. EM calculations
are made in my PC (Intel). For partial and full MD calcultations I
tried to move on cluster runing the same version of GROMACS (3.2.1,
single prec.).

After partial MD start:

   Energies (kJ/mol)
          Angle    Proper Dih.  Improper Dih.          LJ-14     Coulomb-14
    2.97705e+02    3.02067e+02    9.27499e+02    3.75323e+03    3.42362e+01
        LJ (SR)        LJ (LR)   Coulomb (SR)   Coulomb (LR)      Potential
    2.14049e+05   -1.36975e+03   -8.50726e+05            nan            nan
    Kinetic En.   Total Energy    Temperature Pressure (bar)
            nan            nan            nan            nan

Fatal error: ci = -2147483648 should be in 0 .. 4095 [FILE nsgrid.c, LINE 218]

Check EM calculation and reveald that there is no energy data!:

Steepest Descents converged to machine precision in 15 steps,
but did not reach the requested Fmax < 10.
Potential Energy  =            nan
Maximum force     =  1.8116943e+04 on atom 2182
Norm of force     =  1.6091888e+05

ALL the same calculation on another PC (Athlon) runs fine.
Have no idea abou that. I tryed to repeat all steps of job preparation
in cluster, but without any success.

With best wishes
Arturas Ziemys