[gmx-users] why two GTX780ti is slower?

Hello: I've compiled Gromacs in a GPU machine with two GTX780Ti with following command: env CC=icc F77=ifort CXX=icpc CMAKE_PREFIX_PATH=/soft/intel/mkl/include/fftw:/soft/intel/impi/4.1.3.049/intel64:/soft/intel/mkl/lib/intel64 cmake ..

Re: [gmx-users] Positive energy .

Hello Justin and others, Thank you for the response . Here are mode details of the system; ( in the order ; MDP settings, itp and top files) . The system consists of 460 water molecules and 1 CO2 molecule . I am trying to find the free energy of solvation for CO2 in water at various temperatu

Re: [gmx-users] Ion concentration of genion vs charmm-gui

Sorry Mark, I did not mean 201, 000 atoms; I meant 201 potassium atoms vs 232 potassium atoms. Its a small difference, and both methods have the same box size. On Tue, May 13, 2014 at 3:38 PM, Mark Abraham wrote: > Sounds like a problem for a high-school chemistry test ;-) concentration = > qua

Re: [gmx-users] REMD - step 12005: Water molecule starting at atom 7030 can not be settled

Thanks a lot! Berendsen barostat works well with V-rescale using REMD, no crashes observed. On Tue, May 13, 2014 at 2:08 PM, Mark Abraham wrote: > On May 13, 2014 7:21 AM, "Steven Neumann" wrote: > > > > Thanks a lot! So which combination of thermostat/barostat would provide a > > proper ensemb

Re: [gmx-users] quesion about MIC

I recall reading somewhere that MIC support is only native at this point. That is, you can run simulations entirely on MICs, but not hybrid CPU-MIC simulations. Don't recall where I read that and can't easily find it online. Chris. From: gromacs.org_gmx-

[gmx-users] quesion about MIC

Dear All, I am a new user of gromacs, and I want to use gromacs on MIC. Recently I see the gromacs-5.0beta2 published, and in the lists of rebision changes, I see it has been added the SIMD supported for Intel MIC. But when run the gromacs on MIC, it doesn't use the MIC card. On the other hand

Re: [gmx-users] Ion concentration of genion vs charmm-gui

Actually, your target concentration usually refers to the macroscopic solution bulk, not to the simulation box. Therefore, it is generally more correct to think in terms of the ion:solvent molar ratios -- that is, turn your molar concentration into the corresponding macroscopic ion:solvent mola

Re: [gmx-users] Ion concentration of genion vs charmm-gui

Sounds like a problem for a high-school chemistry test ;-) concentration = quantity / volume. From the factor of ~1000, probably you're inputting box sizes in A instead of nm, or vice-versa, somewhere. Mark On Tue, May 13, 2014 at 8:43 PM, Ali Khan wrote: > Hey, > > I generated a membrane in c

Re: [gmx-users] Can't make links on GROMACS

On Tue, May 13, 2014 at 7:56 PM, Ooker wrote: > Justin Lemkul writes: > > > > > > > On 5/13/14, 6:39 AM, Ly Minh Nhat wrote: > > > Justin: why does it remove? > > > > At present, it doesn't really make sense. In 5.0-rc1, all the tools are > simply > > links to a binary called gmx that execute

[gmx-users] Ion concentration of genion vs charmm-gui

Hey, I generated a membrane in charmm-gui that I will be simulating in gromacs. I am using genion to add 0.14 M potassium by using the -conc flag. In a box with dimensions 13.5355 nm x 13.5355 nm x 13 nm, genion states that I should add 201 K atoms. When I constructed the system in charmm-gui, the

Re: [gmx-users] Can't make links on GROMACS

Justin Lemkul writes: > > > On 5/13/14, 6:39 AM, Ly Minh Nhat wrote: > > Justin: why does it remove? > > At present, it doesn't really make sense. In 5.0-rc1, all the tools are simply > links to a binary called gmx that executes all of the programs, so linking to > links is really not con

Re: [gmx-users] Positive energy .

On 5/13/14, 11:21 AM, Rasoul Nasiri wrote: Why do you think positive values are doubtful? What matters are energy differences between different snapshots The absolute energy is meaningless in the FF framework. Not entirely true. While the actual value itself is totally force field-dependen

Re: [gmx-users] Positive energy .

Why do you think positive values are doubtful? What matters are energy differences between different snapshots The absolute energy is meaningless in the FF framework. Rasoul On Tue, May 13, 2014 at 3:59 PM, Bryan Roessler wrote: > Did you hit your nsteps limit? 2 probably isn't enough step

Re: [gmx-users] Positive energy .

Did you hit your nsteps limit? 2 probably isn't enough steps to get below 10 Fmax. *Bryan Roessler | Graduate Research Assistant* UAB | The University of Alabama at Birmingham *uab.edu/cmdb * Knowledge that will change your world On Tue, May 13, 2014 at 8:30 AM, sujithka

Re: [gmx-users] Simulations in NVE ensemble with CHARMM36 - update

On 5/13/14, 10:44 AM, ABEL Stephane 175950 wrote: Hello Justin Thanks for your reply. I have finally found why I obtained a large drift and decrease of the temp of my system during the NVE : I stupidly mixed different cpt and tpr files. If I use the right files and the following parameters b

Re: [gmx-users] Simulations in NVE ensemble with CHARMM36 - update

Hello Justin Thanks for your reply. I have finally found why I obtained a large drift and decrease of the temp of my system during the NVE : I stupidly mixed different cpt and tpr files. If I use the right files and the following parameters below suggested by M. Shirts in [1] for my tests: rl

Re: [gmx-users] g_cluster: treating identical molecules as interchangeable

Hi, If you have a small number of identical molecules you can achieve something along those lines by creating a trajectory where, say, the first half is the coordinates of the first molecule and the second half is the coordinated of the second molecule (for the case of two identical molecules).

Re: [gmx-users] npt dynamics generate a lot of pdb files

But, I modify nstlist only for 500 ps to fix the problem that comming from NVT. After that, I change nstlist to default value for MD. Natalia 2014-05-08 12:51 GMT-04:00 Mark Abraham : > By modifying nstlist, you're very likely fixing a symptom, rather than > fixing the problem. > > Mark > > > On

Re: [gmx-users] g_cluster: treating identical molecules as interchangeable

Hi, I would expect there is no way to do this with g_cluster. g_cluster -h should mention it if it did exist. The only way to implement it is to iterate over all permutations of equivalent molecules. That's doable, but it would have to have been implemented deliberately. Mark On Tue, May 13, 20

[gmx-users] Positive energy .

Dear all, I am running simulations for a water-CO2 system using the OPLSAA forcefield, where I use TIP4P/Ice water model and the EPM2 model for CO2. After energy minimization I found the value of potential energy to be positive, and also the maximum force value was high (I requested Fmax <10 )

Re: [gmx-users] Can't make links on GROMACS

On 5/13/14, 6:39 AM, Ly Minh Nhat wrote: @Justin: why does it remove? At present, it doesn't really make sense. In 5.0-rc1, all the tools are simply links to a binary called gmx that executes all of the programs, so linking to links is really not convenient. @Chandan: I did use sudo whe

Re: [gmx-users] Simulating protein peptide complex

On 5/13/14, 4:30 AM, Ruben Cloete wrote: Hi, Justin regarding protein peptide simulation do i prepare the peptide as i would a ligand (PRODRG) and add topology to .gro file? No, just run it through pdb2gmx. Is there a tutorial for protein protein simulation? No, because it's not really

[gmx-users] g_cluster: treating identical molecules as interchangeable

Hi all, In a simulation of, say, 2 units of the same molecule, it would be useful if g_cluster could treat the two molecules as being interchangeable. For example, if molecule A has three atoms at coordinates C1, C2 and C3, and identical molecule B has three atoms at coordinates C4, C5 and C6, th

Re: [gmx-users] GPU not found

On 13.05.2014 12:42, Albert wrote: Now the problem solved now after I reinstall NVIDIA driver. When I run in single GPU, it works fine. However, when I try to run : mpirun -np 2 mdrun_mpi -s test.tpr it failed with messages: [mpiexec@cudaA.compuBio] HYDU_getfullhostname (./utils/others/others.c

Re: [gmx-users] GPU not found

Now the problem solved now after I reinstall NVIDIA driver. When I run in single GPU, it works fine. However, when I try to run : mpirun -np 2 mdrun_mpi -s test.tpr it failed with messages: [mpiexec@cudaA.compuBio] HYDU_getfullhostname (./utils/others/others.c:136): getaddrinfo error (hostname: c

Re: [gmx-users] Can't make links on GROMACS

@Justin: why does it remove? @Chandan: I did use sudo when make links. Off topic question: if I disable mail delivery but still keep subcribe the mail list, will I receive your answer to my mail? Thank you. On Wed, May 7, 2014 at 7:05 PM, Justin Lemkul wrote: > > > On 5/7/14, 2:04 AM, Ooker w

Re: [gmx-users] GPU not found

On 13.05.2014 12:00, Albert wrote: Hello: thx for reply. it is very strange, I cannot run this command as normal user: Failed to initialize NVML: Insufficient Permissions here is the output I use sudo to run the command you mentioned: Please look into the solutions posted here: http://stackove

Re: [gmx-users] GPU not found

Hello: thx for reply. it is very strange, I cannot run this command as normal user: Failed to initialize NVML: Insufficient Permissions here is the output I use sudo to run the command you mentioned: +--+ | NVIDIA-SMI 331.67 Driver Versi

Re: [gmx-users] GPU not found

On 13.05.2014 11:18, Albert wrote: I compiled Gromacs in a GPU machine with two GTX780Ti with following command: env CC=icc F77=ifort CXX=icpc CMAKE_PREFIX_PATH=/soft/intel/mkl/include/fftw:/soft/intel/impi/ 4.1.3.049/intel64:/soft/intel/mkl/lib/intel64 cmake .. -DBUILD_SHARED_LIB=OFF -DBUILD_TE

[gmx-users] GPU not found

Hello: I compiled Gromacs in a GPU machine with two GTX780Ti with following command: env CC=icc F77=ifort CXX=icpc CMAKE_PREFIX_PATH=/soft/intel/mkl/include/fftw:/soft/intel/impi/ 4.1.3.049/intel64:/soft/intel/mkl/lib/intel64 cmake .. -DBUILD_SHARED_LIB=OFF -DBUILD_TESTING=OFF -DCMAKE_INSTALL_PRE

[gmx-users] Simulating protein peptide complex

Hi, Justin regarding protein peptide simulation do i prepare the peptide as i would a ligand (PRODRG) and add topology to .gro file? Is there a tutorial for protein protein simulation? Hope you can assist Best regards Ruben On 26/04/2014 14:50, Justin Lemkul wrote: On 4/26/14, 8:42 AM, An

[gmx-users] Regarding calculation of energy auto-correlation function

Dear Gromacs User's, Could anybody tell me how to calculate energy auto-correlation function of the solvent from MDS trajectory of a production run. I tried using g_energy_d option but it didn't work properly, maybe I made a mistake in the commands. Following is the

Re: [gmx-users] Compiling gromacs 4.6.5 on cray XT5m

Hello again Problem solved, I think. At least it's compiling now. I downloaded fftw myself and built it with --enable-static --disable-fortran, and that seems to have worked. Thanks /PK 2014-05-13 9:06 GMT+02:00 Gmx QA : > Dear users > > I am having problems compiling gromacs 4.6.5 on a (rath

[gmx-users] topol_pope.top

Dear Sir, If am using POPE model for simulation , where we have the access of p topol_pope.top. regards -- B.Suriyanarayanan -- Gromacs Users mailing list * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting! * Can't post? Read http://www.

Re: [gmx-users] Citing gromacs tools

Thank you for the clarification, Prof. David. On Tuesday, May 13, 2014, David van der Spoel wrote: > On 2014-05-13 07:52, rajat desikan wrote: > >> Hi All, >> In many papers, the authors cite the main gromacs paper and simply declare >> in the methods section that they used the gromacs analysis

[gmx-users] Compiling gromacs 4.6.5 on cray XT5m

Dear users I am having problems compiling gromacs 4.6.5 on a (rather ancient) Cray XT5m machine. My approach has been the following, but it stops during the make stage when processing fftw. 1. Download gromacs-4.6.5 tar ball tar -xvg gromacs-4.6.5.tar cd gromacs-4.6.5 mkdir build cd build Then