[gmx-users] Crashing abnormally
akwsum at gmail.com
Tue Nov 2 01:40:29 CET 2004
For testing pruposes, I have running Tieleman's 128 DPPC bilayer in
parallel (4 processors - 2 dual nodes), and I am getting some strange
results. I started two runs with same configuration. One was set to
run with a double precision version of Gromacs and it crashed after
about 1 ns with:
**Fatal error: ci = 1777 should be in 0 .. 1727 [FILE nsgrid.c, LINE 218]
The other is running with a single precision version of Gromacs and it
continues without problems.
I have seen many of my jobs crashing with this Fatal error several
times. There are similar postings in the mailing list, but the problem
still puzzles me.
I am running Gromacs 3.2.1 compiled with MPICH 1.2.5. The system is an
Opteron cluster. The simulations are set to run in the NPT ensemble
with PME and a 2-fs time-step. There is nothing unusual about the
setting, as far as I can tell.
I would appreciate if anyone could comment on this error and perhaps
provide some input/feedback what may be happening. Is it possible
there is something to do with hardware/memory problem?
More information about the gromacs.org_gmx-users