[gmx-users] strange lincs warning with version 4.6
Dear GROMCS user, I installed the git gromacs VERSION 4.6-dev-20121117-7a330e6-dirty on my local desktop (2*GTX 670 + i7) and everything works as smooth as possible. The outcomes are very reasonable and match the outcome of the 4.5.5 version without GPU acceleration. On our cluster (M2090+2*Xeon X5650) I installed the VERSION 4.6-dev-20121120-0290409. Using the same .tpr file used for runs with my desktop I get lincs warnings that the watermolecules can't be settled. My .mdp file looks like: ; title= ttt cpp = /lib/cpp include = -I../top constraints = hbonds integrator = md cutoff-scheme = verlet ;define = -DPOSRES; for possition restraints dt = 0.002; ps ! nsteps = 1 \ nstcomm = 25; frequency for center of mass motion removal nstcalcenergy = 25 nstxout = 10; frequency for writting the trajectory nstvout = 10; frequency for writting the velocity nstfout = 10; frequency to write forces to output trajectory nstlog = 1; frequency to write the log file nstenergy = 1; frequency to write energies to energy file nstxtcout = 1 xtc_grps= System nstlist = 25; Frequency to update the neighbor list ns_type = grid; Make a grid in the box and only check atoms in neighboring grid cells when constructing a new neighbor rlist = 1.4; cut-off distance for the short-range neighbor list coulombtype = PME; Fast Particle-Mesh Ewald electrostatics rcoulomb= 1.4; cut-off distance for the coulomb field vdwtype = cut-off rvdw= 1.4; cut-off distance for the vdw field fourierspacing = 0.12; The maximum grid spacing for the FFT grid pme_order = 6; Interpolation order for PME optimize_fft= yes pbc= xyz Tcoupl = v-rescale tc-grps = System tau_t = 0.1 ref_t = 300 energygrps = Protein Non-Protein Pcoupl = no;berendsen tau_p = 0.1 compressibility = 4.5e-5 ref_p = 1.0 nstpcouple= 5 refcoord_scaling= all Pcoupltype = isotropic gen_vel = no;yes gen_temp= 300 gen_seed= -1 Thanks a lot Sebastian -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Re,i don't know how can i determine emtol
On 11/22/2012 09:32 PM, Ali Alizadeh wrote: 1- In your opinion, Can i simulate that system? In my (humble) opinion: 1)Of course you can simulate that system...however I doubt that, without starting from the exact initial configuration with the exactly same set-up, you can get the same results (i.e. see the nucleation). The onset of ice nucleation is a random process and requires very long simulation (the paper that you posted was analysing micro-second trajectories!!). There is the risk that you could try several different initial configurations at several temperature without getting anything. However, read carefully that paper, I do not remember all the details. If you are interested in ice crystal growth, I would suggest to start with an initial water/ice system: at temperature below the melting one you will see formation of new ice starting from the initial ice matrix. I do not know how can ice crystal in initial step, i study papers and i know ice , crystals formation and dissociation of crystals is very difficult, and we can use this method(as you said above) for melting or freezing(if i say correct) My problem is this case, how can i have a crystal in initial step, You have three options 1) Surfing the web searching for some structure. 2)Try to implement the proton disordering algorithm of Buch et al., (V. Buch, P. Sandler and J. Sadlej, J. Phys. Chem. B, 1998, 102, 8641–8653) which specifies orientations of water molecules such that ice Ih Bernal–Fowler constraints for each molecule are satisfied. (J. D. Bernal and R. H. Fowler, J. Chem. Phys., 1933, 1, 515–548) 3) Someone kindly offer you his own structure, e.g. http://marge.uochb.cas.cz/~gladich/Teaching.html That's right, Is it possible that i know how can you construct it? You use TIP5P, and i do not know your force field? In fact i should construct this structure by self, Can you give some advices so that i can construct it? I constructed the structure using the method 2) with the Buch's algorithm. Concerning TIP5P and TIP5P-Ew, here below the references: 1) Mahoney and Jorgensen, J. Chem. Phys., 2000, VOLUME 112, NUMBER 20 2) S. W. Rick, J. Chem. Phys., 2004, 120, 6085–6093. All the best Ivan -- -- Ivan Gladich, Ph.D. Postdoctoral Fellow Academy of Sciences of the Czech Republic Institute of Organic Chemistry and Biochemistry AS CR, v.v.i. Flemingovo nám. 2. 166 10 Praha 6 Czech Republic Tel: +420775504164 e-mail: ivan.glad...@uochb.cas.cz web page:http://www.molecular.cz/~gladich/ - -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] Query about GPU version of Gromacs
Hai all, Can any body suggest me which is most stable version of GPU-GROMCS. and as far as Gromacs site is concerned currently all the versions are unstable. If there is no stable version for GPU is available then when would a stable version for GPU would be released. and another query is when i was compiling this gromacs-4.5-GPU-beta2_linux-X86_64.tar.gz (Unstable) GPU version of Gromacs i get path error where it is showing that gcc compiler is not at the given path even though path for all the gcc compilers remains the same. any suggestions would be greatly appreciated. Thank you, -- PruthviRaj Bejugam, Junior Research Fellow, Lab. No.9 (New Building) Computational and Systems Biology Lab, National Centre for Cell Science, Ganeshkhindh, Pune, INDIA 411007 -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] REGARDING DOUBT
dear sir, i have give input that is* ./pdb2gmx_d -f activesite16-22.pdb -water tip3p -ignh* but i face some* errror like There is a dangling bond at at least one of the terminal ends and the force field does not provide terminal entries or files. Edit a .n.tdb and/or .c.tdb fi*le. i have made necessary correction in the .rtp file acccording to my pdb file. with regards S.BOOPATHI -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] REGARDING DOUBT
On 11/23/12 7:41 AM, Subramaniam Boopathi wrote: dear sir, i have give input that is* ./pdb2gmx_d -f activesite16-22.pdb -water tip3p -ignh* but i face some* errror like There is a dangling bond at at least one of the terminal ends and the force field does not provide terminal entries or files. Edit a .n.tdb and/or .c.tdb fi*le. i have made necessary correction in the .rtp file acccording to my pdb file. What changes did you make? Do you have any unnatural amino acids or capping groups? What are your terminal residues? What force field are you using? -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] Energy minimization with walls
Hi all, I am simulating a box filled only with water and having walls at Z=0 and Z=Z_box. I am confused that whether energy minimization process removes the bad contacts between wall atoms and solvent or not ? any kind of help would be really appreciated -- View this message in context: http://gromacs.5086.n6.nabble.com/Energy-minimization-with-walls-tp5003229.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Query about GPU version of Gromacs
Hi, The GPU functionality available in GROMACS 4.5 based on OpenMM is likely to be deprecated because we lack the resources to continue to support it - as you can see from the limited documentation you can see for it on our website. I would advise against attempting to build or use anything other than an official GROMACS release. The good news is that the upcoming GROMACS 4.6 release will have native GPU support for a much larger range of algorithms than have been available previously. Unfortunately, we still can't provide a timeline for that release yet. Portability, reliability, high performance and rapid time-to-market tend to be mutually exclusive, I'm afraid! Mark On Fri, Nov 23, 2012 at 10:53 AM, Pruthvi Bejugam pruthvi.n...@gmail.comwrote: Hai all, Can any body suggest me which is most stable version of GPU-GROMCS. and as far as Gromacs site is concerned currently all the versions are unstable. If there is no stable version for GPU is available then when would a stable version for GPU would be released. and another query is when i was compiling this gromacs-4.5-GPU-beta2_linux-X86_64.tar.gz (Unstable) GPU version of Gromacs i get path error where it is showing that gcc compiler is not at the given path even though path for all the gcc compilers remains the same. any suggestions would be greatly appreciated. Thank you, -- PruthviRaj Bejugam, Junior Research Fellow, Lab. No.9 (New Building) Computational and Systems Biology Lab, National Centre for Cell Science, Ganeshkhindh, Pune, INDIA 411007 -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] how to repeat simulation correctly?
On Thu, Nov 22, 2012 at 10:13 AM, Felipe Pineda, PhD luis.pinedadecas...@lnu.se wrote: Would non-deterministic be correct to characterize the nature of MD as well? There is also deterministic chaos ... An MD simulation is normally deterministic, inasmuch as the inputs and algorithm determine the output, even if the model physics being simulated is not determinstic (e.g. stochastic elements to the integration). One needs to be clear about which aspect is of interest. The simulation is often not reproducible, either, because of run-time effects like dynamic load balancing or differences in hardware/compilers/libraries. And what about the outcome of starting several trajectories from the same equilibrated frame as continuation runs, i.e., using its velocities? Could they be considered independent and used to extract that valuable statistics mentioned in a previous posting? Of course, sampling a coloured ball from a bag of balls, putting it back in without letting go, and taking it back out doesn't create a new sample from the bag. If all the balls were slowly changing colour and you were trying to sample the distribution of colours, then that in-and-out process might be a way to create a new sample, but it depends on the timescales involved... So yes, you could start from the same set of positions and velocities and rely on run-time irreproducibility to introduce differences, and chaos to amplify those over simulation time, in order to reach points from which you could make statistically independent simulations. Empirically, you'll need less simulation time to reach that point if you take active steps to make a significant difference, like changing the velocity of every atom. You need to re-equilibrate each time you perturb, but that's generally cheaper than the alternatives. Mark Felipe On 11/22/2012 10:04 AM, Erik Marklund wrote: Stochastic and chaotic are not identical. Chaotic means that differences in the initial state will grow exponentially over time. Erik 22 nov 2012 kl. 09.52 skrev Felipe Pineda, PhD: Won't this same stochastic nature of MD provide for different, independent trajectories even if restarted from a previous, equilibrated frame even without resetting velocities, i.e., as a continuation run using the velocities recorded in the gro file of the selected snapshot? Felipe On 11/22/2012 12:55 AM, Mark Abraham wrote: Generating velocities from a new random seed is normally regarded as good enough. By the time you equilibrate, the chaotic nature of MD starts to work for you. Mark On Nov 21, 2012 1:04 PM, Felipe Pineda, PhD luis.pinedadecas...@lnu.se wrote: So how would you repeat the (let be it converged) simulation from different starting conditions in order to add that valuable statistics you mention? I think this was Albert's question Felipe On 11/21/2012 12:41 PM, Mark Abraham wrote: If a simulation ensemble doesn't converge reliably over a given time scale, then it's not converged over that time scale. Repeating it from different starting conditions still adds valuable statistics, but can't be a replicate. Independent replicated observations of the same phenomenon allow you to assess how likely it is that your set of observations reflect the underlying phenomenon. The problem in sampling-dependent MD is usually in making an observation (equating a converged simulation with an observation). Mark On Wed, Nov 21, 2012 at 8:12 AM, Albert mailmd2...@gmail.com wrote: hello: I am quite confused on how to repeat our MD in Gromacs. If we started from the same equilibrated .gro file with gen_vel= no in md.mdp, we may get exactly the same results which cannot be treated as reasonable repeated running. However, if we use gen_vel=yes for each round of running, sometimes our simulation may not converged at our simulated time scale and we may get two results with large differences. So I am just wondering how to perform repeated MD in Gromacs in a correct way so that our results can be acceptably repeated? thank you very much. Albert -- -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/**mailman/listinfo/gmx-usershttp://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/** Support/Mailing_Lists/Searchhttp://www.gromacs.org/Support/Mailing_Lists/Searchbefore posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/**Support/Mailing_Listshttp://www.gromacs.org/Support/Mailing_Lists --**- Erik Marklund, PhD Dept. of Cell and Molecular Biology, Uppsala University. Husargatan 3, Box 596,75124 Uppsala, Sweden phone:+46 18 471 6688fax: +46 18 511 755 er...@xray.bmc.uu.se
[gmx-users] On the usage of SD integrator as the thermostat
I use the SD integrator with tau_t = 1.0 ps for all of my work, including proteins in aqueous solution or embedded in a lipid membrane. Any value of tau-t is correct, and none will give you the proper dynamics, but I find that the diffusion of both water and lipids is quite reasonable when using tau_t=1.0 ps. I arrived at 1.0 ps after some help from Berk Hess on this list. I suggest that you search out those old posts. Chris. -- original message -- In manual I've found possibility of the usage of the sd (langeven's dynamics) integrator as the thermostat. It's known that friction coefficient in the Langeven's equations is defined as m/Tau_t. So the high values of tau t can be appropriate for the modeling of the thermostat without t_coupl. Also I know that friction coefficient for such simulation must corresponds to the viscosity of the system. In Gromacs manual I've found that Tau-t= 2.0 ps can be appropriate value for such simulations. Does this value suitable for water-soluble system only ? What Tau_t should I use for modeling of the membrane proteins in the lipid-water environment which has higher viscosity ? Thanks for help, James -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] strange lincs warning with version 4.6
Hi, On Fri, Nov 23, 2012 at 9:40 AM, sebastian sebastian.wa...@physik.uni-freiburg.de wrote: Dear GROMCS user, I installed the git gromacs VERSION 4.6-dev-20121117-7a330e6-dirty on my local desktop Watch out, the dirty version suffix means you have changed something in the source. (2*GTX 670 + i7) and everything works as smooth as possible. The outcomes are very reasonable and match the outcome of the 4.5.5 version without GPU acceleration. On our What does outcome mean? If that means performance, than something is wrong, you should see a considerable performance increase (PME, non-bonded, bondeds have all gotten a lot faster). cluster (M2090+2*Xeon X5650) I installed the VERSION 4.6-dev-20121120-0290409. Using the same .tpr file used for runs with my desktop I get lincs warnings that the watermolecules can't be settled. The group kernels have not stabilized yet and there have been some fixes lately. Could you please the latest version and check again. Additionally, you could try running our regression tests suite ( git.gromacs.org/regressiontests) to see if at least the tests pass with the binaries you compiled. Cheers, -- Szilárd My .mdp file looks like: ; title= ttt cpp = /lib/cpp include = -I../top constraints = hbonds integrator = md cutoff-scheme = verlet ;define = -DPOSRES; for possition restraints dt = 0.002; ps ! nsteps = 1 \ nstcomm = 25; frequency for center of mass motion removal nstcalcenergy = 25 nstxout = 10; frequency for writting the trajectory nstvout = 10; frequency for writting the velocity nstfout = 10; frequency to write forces to output trajectory nstlog = 1; frequency to write the log file nstenergy = 1; frequency to write energies to energy file nstxtcout = 1 xtc_grps= System nstlist = 25; Frequency to update the neighbor list ns_type = grid; Make a grid in the box and only check atoms in neighboring grid cells when constructing a new neighbor rlist = 1.4; cut-off distance for the short-range neighbor list coulombtype = PME; Fast Particle-Mesh Ewald electrostatics rcoulomb= 1.4; cut-off distance for the coulomb field vdwtype = cut-off rvdw= 1.4; cut-off distance for the vdw field fourierspacing = 0.12; The maximum grid spacing for the FFT grid pme_order = 6; Interpolation order for PME optimize_fft= yes pbc= xyz Tcoupl = v-rescale tc-grps = System tau_t = 0.1 ref_t = 300 energygrps = Protein Non-Protein Pcoupl = no;berendsen tau_p = 0.1 compressibility = 4.5e-5 ref_p = 1.0 nstpcouple= 5 refcoord_scaling= all Pcoupltype = isotropic gen_vel = no;yes gen_temp= 300 gen_seed= -1 Thanks a lot Sebastian -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/**mailman/listinfo/gmx-usershttp://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/** Support/Mailing_Lists/Searchhttp://www.gromacs.org/Support/Mailing_Lists/Searchbefore posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/**Support/Mailing_Listshttp://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] Different average H bonds with different g_hbond releases
Hi, we have a .xtc and .tpr file. We were interested in the average number of H-bonds in the last 10ns of a 60ns long trajectory. We analyzed the jobs as g_hbond -f traj1_0-60ns.xtc -s topol.tpr -b 5 -num hbond.xvg. We are displaced by having a different number depending on the g_hbond release. Release 4.5.4 : Average number of hbonds per timeframe 163.620 out of 118112 possible Release 4.5.5 : Average number of hbonds per timeframe 168.168 out of 118112 possible Looking at the hbond.xvg file, the number of H-bonds in each frame are clearly different between the two releases. How is this possible ? We checked single versus double precision g_hbonds, same behavior. We checked that the initial part of the output, i.e. all the various g_hbond defaults, they are the same. We tested different computers and compilations, same behavior. The topology and the md run were done with release 4.5.4 if this could be a relevant information. Thanks, Luigi -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Different average H bonds with different g_hbond releases
On 11/23/12 5:23 PM, Luigi CAVALLO wrote: Hi, we have a .xtc and .tpr file. We were interested in the average number of H-bonds in the last 10ns of a 60ns long trajectory. We analyzed the jobs as g_hbond -f traj1_0-60ns.xtc -s topol.tpr -b 5 -num hbond.xvg. We are displaced by having a different number depending on the g_hbond release. Release 4.5.4 : Average number of hbonds per timeframe 163.620 out of 118112 possible Release 4.5.5 : Average number of hbonds per timeframe 168.168 out of 118112 possible Looking at the hbond.xvg file, the number of H-bonds in each frame are clearly different between the two releases. How is this possible ? We checked single versus double precision g_hbonds, same behavior. We checked that the initial part of the output, i.e. all the various g_hbond defaults, they are the same. We tested different computers and compilations, same behavior. The topology and the md run were done with release 4.5.4 if this could be a relevant information. There was a bug that was fixed in May 2011 wherein 4.5.4 reported too few hydrogen bonds. commit 91a481fad7ef0d87a4f8b2cb633c9dc40644350c Author: Erik Marklund er...@anfinsen.bmc.uu.se Date: Tue May 10 14:37:10 2011 +0200 Fixed long standing bug where the merging resulted in too few hbonds. -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Gromacs 4.6 segmentation fault with mdrun
Hi Szilárd and Roland, Thanks for the clear explanation! I will compile release-4.6 (instead of the nbnxn_hybrid_acc branch) and do some further testing in a few weeks since I'm currently using the machine for production-runs with gmx-4.5.5. Thanks for your time and effort! regards, raf On Thu, 2012-11-22 at 00:23 +0100, Szilárd Páll wrote: Roland, He explicitly stated that he is using 20da718 which is also from the nbnxn_hybrid_acc branch. Raf, as Roland said, get the release-4-6 ad try again! There's an important thing to mention: your hardware configuration is probably quite imbalanced and the default settings are certainly not the best to run with: two MPI processes/threads with 24 OpenMP threads + a GPU each. GROMACS works best with balanced hardware configuration and yours is certainly not balanced, the GPUs will not be able to keep up with 64 CPU cores. Regarding the run configuration most importantly, in most cases you should avoid running a group of OpenMP threads across sockets (except on Intel, =12-16 threads). On these Opterons running OpenMP at most on a half CPU is recommended (the CPUs are in reality two CPU dies bolted together) and in fact you might be better off with even less threads per MPI process/thread. This means that multiple processes will have to share a GPU which is not optimal and work only with MPI in the current version. So to conclude, to get the best performance you should try a few combinations: # process 0,1 will use GPU0, process 2,3 GPU1 # this avoids running across sockets, but for aforementioned reasons it will still be suboptimal mpirun -np 4 mdrun_mpi -gpu_id 0011 # process 0,1,2,3 will use GPU0, process 4,5,6,7 GPU1 # this config will probably still be slower than the next one mpirun -np 8 mdrun_mpi -gpu_id 1 # process 0,1,2,3,4,5,6,7 will use GPU0, process 8,9,10,11,12,13,14,15 GPU1 # this config will probably still be slower than the next one mpirun -np 16 mdrun_mpi -gpu_id 1 You should go ahead and try with 32 and 64 processes as well, I suspect that 2 or 3 threads/process will be the fastest. Depending on what system you are simulating, this could lead to load imbalance, but that you'll have to see. If it turns out that the Wait for GPU time is more than a few percent (which will probably be the case), it means that a GTX 580 is not fast enough for two of these Opterons. What you can try is to run using the hybrid mode with -nb gpu_cpu which might help. Cheers, -- Szilárd On Sat, Nov 17, 2012 at 3:11 AM, Roland Schulz rol...@utk.edu wrote: Hi Raf, which version of Gromacs did you use? If you used branch nbnxn_hybrid_acc please use branch release-4-6 instead and see whether that fixes your issue. If not please open a bug and upload your log file and your tpr. Roland On Thu, Nov 15, 2012 at 5:13 PM, Raf Ponsaerts raf.ponsae...@med.kuleuven.be wrote: Hi Szilárd, I assume I get the same segmentation fault error as Sebastian (don't shoot if not so). I have 2 NVIDA GTX580 cards (and 4x12-core amd64 opteron 6174). in brief : Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7fffc07f8700 (LWP 32035)] 0x761de301 in nbnxn_make_pairlist.omp_fn.2 () from /usr/local/gromacs/bin/../lib/libmd.so.6 Also -nb cpu with Verlet cutoff-scheme results in this error... gcc 4.4.5 (Debian 4.4.5-8), Linux kernel 3.1.1 CMake 2.8.7 If I attach the mdrun.debug output file to this mail, the mail to the list gets bounced by the mailserver (because mdrun.debug 50 Kb). Hoping this might help, regards, raf === compiled code : commit 20da7188b18722adcd53088ec30e5f256af62f20 Author: Szilard Pall pszil...@cbr.su.se Date: Tue Oct 2 00:29:33 2012 +0200 === (gdb) exec mdrun (gdb) run -debug 1 -v -s test.tpr Reading file test.tpr, VERSION 4.6-dev-20121002-20da718 (single precision) [New Thread 0x73844700 (LWP 31986)] [Thread 0x73844700 (LWP 31986) exited] [New Thread 0x73844700 (LWP 31987)] [Thread 0x73844700 (LWP 31987) exited] Changing nstlist from 10 to 50, rlist from 2 to 2.156 Starting 2 tMPI threads [New Thread 0x73844700 (LWP 31992)] Using 2 MPI threads Using 24 OpenMP threads per tMPI thread 2 GPUs
Re: [gmx-users] Different average H bonds with different g_hbond releases
If we got the results by 4.5.4, what will be the method to analyze it by 4.5.5? By a pathch or by installation of 4.5.5 to analyze the 4.5.4 results? Cheers, Acoot --- On Sat, 24/11/12, Justin Lemkul jalem...@vt.edu wrote: From: Justin Lemkul jalem...@vt.edu Subject: Re: [gmx-users] Different average H bonds with different g_hbond releases To: Discussion list for GROMACS users gmx-users@gromacs.org Received: Saturday, 24 November, 2012, 9:30 AM On 11/23/12 5:23 PM, Luigi CAVALLO wrote: Hi, we have a .xtc and .tpr file. We were interested in the average number of H-bonds in the last 10ns of a 60ns long trajectory. We analyzed the jobs as g_hbond -f traj1_0-60ns.xtc -s topol.tpr -b 5 -num hbond.xvg. We are displaced by having a different number depending on the g_hbond release. Release 4.5.4 : Average number of hbonds per timeframe 163.620 out of 118112 possible Release 4.5.5 : Average number of hbonds per timeframe 168.168 out of 118112 possible Looking at the hbond.xvg file, the number of H-bonds in each frame are clearly different between the two releases. How is this possible ? We checked single versus double precision g_hbonds, same behavior. We checked that the initial part of the output, i.e. all the various g_hbond defaults, they are the same. We tested different computers and compilations, same behavior. The topology and the md run were done with release 4.5.4 if this could be a relevant information. There was a bug that was fixed in May 2011 wherein 4.5.4 reported too few hydrogen bonds. commit 91a481fad7ef0d87a4f8b2cb633c9dc40644350c Author: Erik Marklund er...@anfinsen.bmc.uu.se Date: Tue May 10 14:37:10 2011 +0200 Fixed long standing bug where the merging resulted in too few hbonds. -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing list gmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Different average H bonds with different g_hbond releases
On 2012-11-24 05:41, Acoot Brett wrote: If we got the results by 4.5.4, what will be the method to analyze it by 4.5.5? By a pathch or by installation of 4.5.5 to analyze the 4.5.4 results? In practice there is no problem to have a number of gromacs versions installed. It is typically not recommended to switch gromacs versions of mdrun during a project - unless there are know issues - but for the analysis this is less critical. In this case 4.5.5 should be used. Cheers, Acoot --- On Sat, 24/11/12, Justin Lemkul jalem...@vt.edu wrote: From: Justin Lemkul jalem...@vt.edu Subject: Re: [gmx-users] Different average H bonds with different g_hbond releases To: Discussion list for GROMACS users gmx-users@gromacs.org Received: Saturday, 24 November, 2012, 9:30 AM On 11/23/12 5:23 PM, Luigi CAVALLO wrote: Hi, we have a .xtc and .tpr file. We were interested in the average number of H-bonds in the last 10ns of a 60ns long trajectory. We analyzed the jobs as g_hbond -f traj1_0-60ns.xtc -s topol.tpr -b 5 -num hbond.xvg. We are displaced by having a different number depending on the g_hbond release. Release 4.5.4 : Average number of hbonds per timeframe 163.620 out of 118112 possible Release 4.5.5 : Average number of hbonds per timeframe 168.168 out of 118112 possible Looking at the hbond.xvg file, the number of H-bonds in each frame are clearly different between the two releases. How is this possible ? We checked single versus double precision g_hbonds, same behavior. We checked that the initial part of the output, i.e. all the various g_hbond defaults, they are the same. We tested different computers and compilations, same behavior. The topology and the md run were done with release 4.5.4 if this could be a relevant information. There was a bug that was fixed in May 2011 wherein 4.5.4 reported too few hydrogen bonds. commit 91a481fad7ef0d87a4f8b2cb633c9dc40644350c Author: Erik Marklund er...@anfinsen.bmc.uu.se Date: Tue May 10 14:37:10 2011 +0200 Fixed long standing bug where the merging resulted in too few hbonds. -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- David van der Spoel, Ph.D., Professor of Biology Dept. of Cell Molec. Biol., Uppsala University. Box 596, 75124 Uppsala, Sweden. Phone: +46184714205. sp...@xray.bmc.uu.sehttp://folding.bmc.uu.se -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists