[gmx-users] no-pbc simulation in parallel

[Vitaly Chaban]

Hi all,

I am experiencing a strange error while trying to perform a parallel
run with pbc=no. Just a core.*** file (at the very beginning) is
created and nothing is written into the logfile. The simulation just
stops. I use mpirun -np 8 mdrun_407f -pd 2 2 2 .

The same system runs correctly if PBC=XYZ. The system is rather big -
10x10x10nm with ~40 000 atoms

My parameters are below:
==========================
integrator               = md
dt                       = 0.0025
nstxout                  = 0
nstvout                  = 0
nstfout                  = 0
emtol                    = 100
nsteps                   = 50000
nstlog                   = 5000
nstenergy                = 1000
nstxtcout                = 1000
xtc-precision            = 1000
nstlist                  = 5
comm_mode = ANGULAR
ns_type                  = simple
pbc                      = no
rlist                    = 1.4
coulombtype              = cut-off
rcoulomb                 = 1.4
vdw-type                 = shift
rvdw-switch              = 1.2
rvdw                     = 1.3
DispCorr                 = no
table-extension          = 1
Tcoupl                   = V-rescale
tc-grps                  = System
tau_t                    = 0.1
ref_t                    = 373
;Pcoupl                   = no ;Parrinello-Rahman
;Pcoupltype               = isotropic
;tau_p                    = 0.5
;compressibility          = 4.5e-5
;ref_p                    = 100
==========================

Is it a bug or what?

Thank you.

Dr. Vitaly Chaban