[gmx-users] do remd in the npt ensemble:Warning: pressure scaling more than 1%, mu: 333821 333821 333821

[Mark Abraham]

On 24/11/2011 6:31 PM, 杜波 wrote:
> dear teacher,
> when i do remd in the npt ensemble.
>
> =========md.mdp===================
> ; Start time and timestep in ps
> tinit = 0
> dt = 0.000001
> nsteps = 50000000
> ; For exact run continuation or redoing part of a run
>
> ; Temperature coupling
> Tcoupl = nose-hoover
> ; Groups to couple separately
> tc_grps = system
> ; Time constant (ps) and reference temperature (K)
> tau_t = 0.3
> ref_t = 800
> ; Pressure coupling
> Pcoupl = berendsen
> pcoupltype = isotropic
> ; Time constant (ps), compressibility (1/bar) and reference P (bar)
> tau_p = 0.5
> compressibility = 6.5e-5
> ref_p = 1.0
> ==========error========================
> step 999900, will finish Mon Nov 28 00:11:22 2011
> step 1000000, will finish Mon Nov 28 00:11:19 2011
>
> Step 1000002 Warning: pressure scaling more than 1%, mu: 333821 333821 
> 333821
>
> Step 1000002 Warning: pressure scaling more than 1%, mu: 455440 455440 
> 455440
>
> -------------------------------------------------------
> Program mdrun_mpi_4.5.5, VERSION 4.5.5
> Source code file: smalloc.c, line: 214
>
> Fatal error:
> Not enough memory. Failed to realloc -5476105916 bytes for grid->nra, 
> grid->nra=0x38ce0
> (called from file nsgrid.c, line 483)
> For more information and tips for troubleshooting, please check the 
> GROMACS
> website at http://www.gromacs.org/Documentation/Errors
> -------------------------------------------------------
> ===============command=====================
> mpirun n8,8,8,11,11,11,11,2 -np $n_thread 
> /export/software/bin/mdrun_mpi_4.5.5 -multidir ./0/ ./1/ ./2/ ./3/ 
> ./4/ ./5/ ./6/ ./7/ ./8/ ./9/ ./10/ ./11/ ./12/ ./13/ ./14/ ./15/ 
> -replex 500000 -nice 0 -s pmma.tpr -o md -c after_md -pd -v >& log1 &
>
> how could i do except use the smaller dt or big/small tau_p ?
>

Surprisingly, the URL provided in the error message contains a useful 
source of information:
http://www.gromacs.org/Documentation/Errors#Pressure_scaling_more_than_1.25

Mark