[gmx-users] gcc problem
David van der Spoel
spoel at xray.bmc.uu.se
Mon Nov 5 09:03:13 CET 2007
余翔 wrote:
> Dear all:
>
> I have a problem when I was running mdrun on different computers. I made a .tpr file including coordinate and velocity of each atom from a crash of mdrun and upload it to four computers whose distribution version of linux and gcc were as follows:
> computer A: Redhat 9.0, gcc 3.2.2
> computer B: Turbo Linux 10,gcc 3.4.3
> computer C: Redhat 7.2, gcc 2.96
> computer D: Fedora core 6, gcc 3.4.6(the default version is 4.1.1, so I reinstalled an old version on it)
> All the gmx are compiled in float and mpi mode.
> The results of mdrun (the final .gro file, the trajectory file and the energy file) that start from the same .tpr file obtained on these computers are totally different, but if I ran mdrun on one computer from this .tpr file twice or more times, the results are all the same. So I don't know why the "mdrun" compiled by different versions of gcc would give such different results and which result should I trust. Does anybody know why it happens? I beg for your help...
>
> this is the content of my .mdp file, I have modified nstlist, ns_type, pbc, rlist, rvdw and rcoulomb for many times, but it does not work.
>
It is because MD is chaotic, and small differences in compilation, e.g.
due to optimization will change the trajectory. However, your averages
should be the same. Check potential energy after equilibration, and its
fluctuations as well.
> title =
> cpp = /usr/bin/cpp
> include =
> define = -DFLEXIBLE
>
> integrator = md
> tinit = 0
> dt = 0.002
> nsteps = 12000
> init_step = 0
> comm-mode = Angular
> nstcomm = 1
> comm-grps =
>
> bd-fric = 0
> ld-seed = 1993
>
> emtol = 100
> emstep = 0.01
> niter = 20
> fcstep = 0
> nstcgsteep = 1000
> nbfgscorr = 10
>
> nstxout = 100
> nstvout = 1000
> nstfout = 0
> nstcheckpoint = 1000
> nstlog = 1000
> nstenergy = 100
> nstxtcout = 0
> xtc-precision = 1000
> xtc-grps =
> energygrps =
>
> nstlist = 1
> ns_type = grid
> pbc = xyz
> rlist = 0.9
> domain-decomposition = no
>
> coulombtype = Cut-off
> rcoulomb-switch = 0
> rcoulomb = 1.2
> epsilon_r = 1
> epsilon_rf = 1
> vdw-type = Cut-off
> rvdw-switch = 0
> rvdw = 0.9
> DispCorr = No
> table-extension = 1
> energygrp_table =
> fourierspacing = 0.12
> fourier_nx = 0
> fourier_ny = 0
> fourier_nz = 0
> pme_order = 4
> ewald_rtol = 1e-05
> ewald_geometry = 3d
> epsilon_surface = 0
> optimize_fft = no
>
> gb_algorithm = Still
> nstgbradii = 1
> rgbradii = 2
> gb_saltconc = 0
>
> implicit_solvent = No
>
> Tcoupl = nose-hoover
> tc-grps = System
> tau-t = 0.1
> ref-t = 600
> Pcoupl = no
> Pcoupltype = Isotropic
> tau-p = 1
> compressibility =
> ref-p =
> andersen_seed = 815131
>
> QMMM = no
> QMMM-grps =
> QMmethod =
> QMMMscheme = normal
> QMbasis =
> QMcharge =
> QMmult =
> SH =
> CASorbitals =
> CASelectrons =
> SAon =
> SAoff =
> SAsteps =
> MMChargeScaleFactor = 1
> bOPT =
> bTS =
>
> annealing = no
> annealing_npoints = 11
> annealing_time =
> annealing_temp =
>
> gen_vel = no
> gen-temp = 600
> gen-seed = 173529
>
> constraints = none
> constraint-algorithm = Lincs
> unconstrained-start = no
> Shake-SOR = no
> shake-tol = 0.0001
> lincs-order = 4
> lincs-iter = 1
> lincs-warnangle = 30
> morse = no
>
> energygrp_excl =
>
> disre = No
> disre-weighting = Conservative
> disre-mixed = no
> disre-fc = 1000
> disre-tau = 0
> nstdisreout = 100
> orire = no
> orire-fc = 0
> orire-tau = 0
> orire-fitgrp =
> nstorireout = 100
> dihre = No
> dihre-fc = 1000
> dihre-tau = 0
> nstdihreout = 100
>
> free-energy = no
> init-lambda = 0
> delta-lambda = 0
> sc-alpha = 0
> sc-power = 0
> sc-sigma = 0.3
>
> ; Non-equilibrium MD stuff
> acc-grps =
> accelerate =
> freezegrps =
> freezedim =
> cos-acceleration = 0
> deform =
>
> ; Electric fields
> ; Format is number of terms (int) and for all terms an amplitude (real)
> ; and a phase angle (real)
> E-x =
> E-xt =
> E-y =
> E-yt =
> E-z =
> E-zt =
>
> ; User defined thingies
> user1-grps =
> user2-grps =
> userint1 = 0
> userint2 = 0
> userint3 = 0
> userint4 = 0
> userreal1 = 0
> userreal2 = 0
> userreal3 = 0
> userreal4 = 0
>
>
> YuXiang
> yux at iccas.ac.cn
> 2007-11-05
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> gmx-users mailing list gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
--
David van der Spoel, Ph.D.
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone: +46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se spoel at gromacs.org http://folding.bmc.uu.se
More information about the gromacs.org_gmx-users
mailing list