[gmx-users] the crashed run
hi GMX users i use gromacs 4.6.3 double precision for protein ligand complex during the 20 ns. my run crashed, I'm not sure that restart my run from cpt file or run a new production. is the result of started again run a reliable like a non crashed run. thanks for your help -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] the crashed run
_prev.cpt might be ok. I think it writes those in case the current cpt causes problems On 8 Dec 2013, at 09:57, Mahboobeh Eslami wrote: > hi GMX users > i use gromacs 4.6.3 double precision for protein ligand complex during the 20 > ns. my run crashed, I'm not sure that restart my run from cpt file or run a > new production. > is the result of started again run a reliable like a non crashed run. > > thanks for your help > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a > mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] charge correction in topology file from PRODRG
For gromos ff you can also use the ATB server. From Alan Mark on Brisbane. It combine typography and non-bonded parameters all at once. It is not perfect but pretty good. > On Dec 8, 2013, at 8:27, Mahboobeh Eslami wrote: > > hi all my friends > I use PRODRG and antechmber for building topology and coordinate files for my > ligand separately. > i want to use GROMOS force field so i must to use the topology from PRODRG > server. can i use the topology of antechamber for charge correction in > topology fiel from PRODRG . > In general, are special principlesessential for charge correction in topology > file from PRODRG > thanks for your help > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a > mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] Fw: the crashed run
thanks for your reply please suggest the best command for restart a crashed run. i use following command mdrun -v -deffnm md -cpi md is this command good or not? thanks On Sunday, December 8, 2013 2:18 PM, "jkrie...@mrc-lmb.cam.ac.uk" wrote: _prev.cpt might be ok. I think it writes those in case the current cpt causes problems On 8 Dec 2013, at 09:57, Mahboobeh Eslami wrote: > hi GMX users > i use gromacs 4.6.3 double precision for protein ligand complex during the 20 ns. my run crashed, I'm not sure that restart my run from cpt file or run a new production. > is the result of started again run a reliable like a non crashed run. > > thanks for your help > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a > mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Charge correction PME
Thanks for your help. Djamila Benrezkallah -- View this message in context: http://gromacs.5086.x6.nabble.com/Charge-correction-PME-tp5012941p5013185.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] load imbalance in multiple GPU simulations
Hi all, My conventional MD run (equilibration) of a protein in TIP3 water had the "Average load imbalance: 59.4 %" when running with 3 GPUs + 12 CPU cores. So I wonder how to tweak parameters to optimize the performance. End of the log file reads: .. M E G A - F L O P S A C C O U N T I N G NB=Group-cutoff nonbonded kernelsNxN=N-by-N cluster Verlet kernels RF=Reaction-Field VdW=Van der Waals QSTab=quadratic-spline table W3=SPC/TIP3p W4=TIP4p (single or pairs) V&F=Potential and force V=Potential only F=Force only Computing: M-Number M-Flops % Flops - Pair Search distance check 78483.330336706349.973 0.1 NxN QSTab Elec. + VdW [F] 11321254.234368 464171423.60995.1 NxN QSTab Elec. + VdW [V&F] 114522.922048 6756852.401 1.4 1,4 nonbonded interactions1645.932918148133.963 0.0 Calc Weights 25454.159073916349.727 0.2 Spread Q Bspline543022.060224 1086044.120 0.2 Gather F Bspline543022.060224 3258132.361 0.7 3D-FFT 1138719.444112 9109755.553 1.9 Solve PME 353.129616 22600.295 0.0 Reset In Box 424.2275001272.682 0.0 CG-CoM 424.3971911273.192 0.0 Bonds 330.706614 19511.690 0.0 Angles1144.322886192246.245 0.0 Propers 1718.934378393635.973 0.1 Impropers 134.502690 27976.560 0.0 Pos. Restr.321.706434 16085.322 0.0 Virial 424.7348267645.227 0.0 Stop-CM 85.184882 851.849 0.0 P-Coupling8484.719691 50908.318 0.0 Calc-Ekin 848.794382 22917.448 0.0 Lincs 313.720420 18823.225 0.0 Lincs-Mat 1564.1465766256.586 0.0 Constraint-V 8651.865815 69214.927 0.0 Constraint-Vir 417.065668 10009.576 0.0 Settle2674.808325 863963.089 0.2 - Total 487878233.910 100.0 - D O M A I N D E C O M P O S I T I O N S T A T I S T I C S av. #atoms communicated per step for force: 2 x 63413.7 av. #atoms communicated per step for LINCS: 2 x 3922.5 Average load imbalance: 59.4 % Part of the total run time spent waiting due to load imbalance: 5.0 % R E A L C Y C L E A N D T I M E A C C O U N T I N G Computing: Nodes Th. Count Wall t (s) G-Cycles % - Domain decomp. 34 2500 42.792 1300.947 4.4 DD comm. load 34 31 0.0000.014 0.0 Neighbor search34 2501 33.076 1005.542 3.4 Launch GPU ops.34 12 6.537 198.739 0.7 Comm. coord. 34 47500 20.349 618.652 2.1 Force 34 50001 75.093 2282.944 7.8 Wait + Comm. F 34 50001 24.850 755.482 2.6 PME mesh 34 50001 597.92518177.76062.0 Wait GPU nonlocal 34 50001 9.862 299.813 1.0 Wait GPU local 34 50001 0.2627.968 0.0 NB X/F buffer ops. 34 195002 33.578 1020.833 3.5 Write traj.34 12 0.506 15.385 0.1 Update 34 50001 23.243 706.611 2.4 Constraints34 50001 70.972 2157.657 7.4 Comm. energies 34 2501 0.386 11.724 0.0 Rest 3 24.466 743.803 2.5 - Total 3 963.89929303.873 100.0 - - PME redist. X/F34 12 121.844 3704.21412.6 PME spread/gather 34 12 300.759 9143.48631.2 PME 3D-FFT 34 12 11
Re: [gmx-users] Fwd: How can i run my system successfully?
Dear friend very thanks for your answers There is one important thing that i must say, after minimization energy i run my system for nvt equilibration in " 500 ps " and run done without any problem , then i applied the output file as input file for nvt equilibration in "15 ns" , but during first "1 ns" run exit: Step 79777, time 159.554 (ps) LINCS WARNING relative constraint deviation after LINCS: rms 0.346309, max 1.958716 (between atoms 13 and 15) bonds that rotated more than 30 degrees: atom 1 atom 2 angle previous, current, constraint length 13 15 90.00.1000 0.2959 0.1000 Wrote pdb files with previous and current coordinates so ,what is your suggestion for solving this problem? thanks in advance bahar On Sat, Dec 7, 2013 at 4:15 PM, Justin Lemkul wrote: > > > On 12/7/13 4:51 AM, bahareh khanoom wrote: > >> Dear friend >> thanks for your answer >> >> first i generate" DRG.itp" for adsorbed molecules, by PRODRG server. >> next i optimize adsorbed moleculs by gaussian >> b3lyp/6-311++g(d) opt Pop=ChelpG >> and the end ,i replaced the charges in" DRG.itp" with charges that >> produced by gaussian. >> >> > How do these charges compare with existing charges for similar functional > groups in the force field? AFAIK, there is no hard evidence as to which QM > method will give the best results from Gromos96 force fields. > > > so ,what is the reason? >> >> > It is hard to say at this point, but one solution is to simulate each > component individually to verify that their topologies are correct and that > your .mdp settings are appropriate (though they look reasonable on first > glance). > > See also http://www.gromacs.org/Documentation/Terminology/ > Blowing_Up#Diagnosing_an_Unstable_System. > > > -Justin > > -- > == > > Justin A. Lemkul, Ph.D. > Postdoctoral Fellow > > Department of Pharmaceutical Sciences > School of Pharmacy > Health Sciences Facility II, Room 601 > University of Maryland, Baltimore > 20 Penn St. > Baltimore, MD 21201 > > jalem...@outerbanks.umaryland.edu | (410) 706-7441 > > == > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] When crystal dimension is less than rcoulomb and rlist
Hello everyone, I am trying to compute the energy of a crystal slab using steepest decent. Learning from previous questions in the group. I am using pbc in all direction. I need to compute this energy for one layer of crystal. Since crystal unit cell vectors (a b and c) can be less than rcoulomb and rlist ... it is obvious that I receive error while generating the tpr file. as an example, lets assume that a=0.7, b=.9 and c=1.1 angstrom, this means if I have a crystal slab of face(100), and b and c are in XY plane, then box dimension along z axes which is equal to 0.7 ... is less than rlist ... how can i compute the energy of a crystal slab then? If I decrease rcoulomb and rlist ... I am introducing artifact in computing electrostatic interaction and obviously, I can not increase box dimension along Z because in that way, crystal wont be periodic along Z anymore. Best and thanks for you useful comments in advance G. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] When crystal dimension is less than rcoulomb and rlist
On 2013-12-08 21:12, Golshan Hejazi wrote: Hello everyone, I am trying to compute the energy of a crystal slab using steepest decent. Learning from previous questions in the group. I am using pbc in all direction. I need to compute this energy for one layer of crystal. Since crystal unit cell vectors (a b and c) can be less than rcoulomb and rlist ... it is obvious that I receive error while generating the tpr file. as an example, lets assume that a=0.7, b=.9 and c=1.1 angstrom, this means if I have a crystal slab of face(100), and b and c are in XY plane, then box dimension along z axes which is equal to 0.7 ... is less than rlist ... how can i compute the energy of a crystal slab then? If I decrease rcoulomb and rlist ... I am introducing artifact in computing electrostatic interaction and obviously, I can not increase box dimension along Z because in that way, crystal wont be periodic along Z anymore. Best and thanks for you useful comments in advance G. Use multiple layers and divide by the number of layers. -- 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 -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] charge correction in topology file from PRODRG
On 12/8/13 6:18 AM, XAvier Periole wrote: For gromos ff you can also use the ATB server. From Alan Mark on Brisbane. It combine typography and non-bonded parameters all at once. It is not perfect but pretty good. For those wondering about some potential implications of topology errors and how to start going about fixing them, as well as an overview of some common QM calculations one can do to try to calculate charges for new groups: http://pubs.acs.org/doi/abs/10.1021/ci100335w -Justin On Dec 8, 2013, at 8:27, Mahboobeh Eslami wrote: hi all my friends I use PRODRG and antechmber for building topology and coordinate files for my ligand separately. i want to use GROMOS force field so i must to use the topology from PRODRG server. can i use the topology of antechamber for charge correction in topology fiel from PRODRG . In general, are special principlesessential for charge correction in topology file from PRODRG thanks for your help -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Fw: the crashed run
On 12/8/13 6:59 AM, Mahboobeh Eslami wrote: thanks for your reply please suggest the best command for restart a crashed run. i use following command mdrun -v -deffnm md -cpi md is this command good or not? If the desired .cpt file is md.cpt, then mdrun -deffnm md -cpi suffices. If you need a previous .cpt file (i.e. md_prev.cpt), then you must specify its name explicitly. Depending on why the run crashed, simply restarting may not be worthwhile. If the crash was due to the system blowing up, it is a waste of time and you should investigate the crash. If it was a hardware failure, disk error, etc. then continuing from a checkpoint is fine. -Justin thanks On Sunday, December 8, 2013 2:18 PM, "jkrie...@mrc-lmb.cam.ac.uk" wrote: _prev.cpt might be ok. I think it writes those in case the current cpt causes problems On 8 Dec 2013, at 09:57, Mahboobeh Eslami wrote: hi GMX users i use gromacs 4.6.3 double precision for protein ligand complex during the 20 ns. my run crashed, I'm not sure that restart my run from cpt file or run a new production. is the result of started again run a reliable like a non crashed run. thanks for your help -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Fwd: How can i run my system successfully?
On 12/8/13 2:38 PM, bahareh khanoom wrote: Dear friend very thanks for your answers There is one important thing that i must say, after minimization energy i run my system for nvt equilibration in " 500 ps " and run done without any problem , then i applied the output file as input file for nvt equilibration in "15 ns" , but during first "1 ns" run exit: Step 79777, time 159.554 (ps) LINCS WARNING relative constraint deviation after LINCS: rms 0.346309, max 1.958716 (between atoms 13 and 15) bonds that rotated more than 30 degrees: atom 1 atom 2 angle previous, current, constraint length 13 15 90.00.1000 0.2959 0.1000 Wrote pdb files with previous and current coordinates so ,what is your suggestion for solving this problem? The same as I suggested in my previous message. You need to (scientifically!) diagnose the possible sources of error. -Justin thanks in advance bahar On Sat, Dec 7, 2013 at 4:15 PM, Justin Lemkul wrote: On 12/7/13 4:51 AM, bahareh khanoom wrote: Dear friend thanks for your answer first i generate" DRG.itp" for adsorbed molecules, by PRODRG server. next i optimize adsorbed moleculs by gaussian b3lyp/6-311++g(d) opt Pop=ChelpG and the end ,i replaced the charges in" DRG.itp" with charges that produced by gaussian. How do these charges compare with existing charges for similar functional groups in the force field? AFAIK, there is no hard evidence as to which QM method will give the best results from Gromos96 force fields. so ,what is the reason? It is hard to say at this point, but one solution is to simulate each component individually to verify that their topologies are correct and that your .mdp settings are appropriate (though they look reasonable on first glance). See also http://www.gromacs.org/Documentation/Terminology/ Blowing_Up#Diagnosing_an_Unstable_System. -Justin -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] load imbalance in multiple GPU simulations
Hi, That's unfortunate, but not unexpected. You are getting a 3x1x1 decomposition where the "middle" cell has most of the protein, hence most of the bonded forces to calculate, while the ones on the side have little (or none). Currently, the only thing you can do is to try using more domains, perhaps with manual decomposition (such that the initial domains will contain as much protein as possible). This may not help much, though. In extreme cases (e.g. small system), even using only two of the three GPUs could improve performance. Cheers, -- Szilárd On Sun, Dec 8, 2013 at 8:10 PM, yunshi11 . wrote: > Hi all, > > My conventional MD run (equilibration) of a protein in TIP3 water had the > "Average load imbalance: 59.4 %" when running with 3 GPUs + 12 CPU cores. > So I wonder how to tweak parameters to optimize the performance. > > End of the log file reads: > > .. > M E G A - F L O P S A C C O U N T I N G > > NB=Group-cutoff nonbonded kernelsNxN=N-by-N cluster Verlet kernels > RF=Reaction-Field VdW=Van der Waals QSTab=quadratic-spline table > W3=SPC/TIP3p W4=TIP4p (single or pairs) > V&F=Potential and force V=Potential only F=Force only > > Computing: M-Number M-Flops % Flops > - > Pair Search distance check 78483.330336706349.973 0.1 > NxN QSTab Elec. + VdW [F] 11321254.234368 464171423.60995.1 > NxN QSTab Elec. + VdW [V&F] 114522.922048 6756852.401 1.4 > 1,4 nonbonded interactions1645.932918148133.963 0.0 > Calc Weights 25454.159073916349.727 0.2 > Spread Q Bspline543022.060224 1086044.120 0.2 > Gather F Bspline543022.060224 3258132.361 0.7 > 3D-FFT 1138719.444112 9109755.553 1.9 > Solve PME 353.129616 22600.295 0.0 > Reset In Box 424.2275001272.682 0.0 > CG-CoM 424.3971911273.192 0.0 > Bonds 330.706614 19511.690 0.0 > Angles1144.322886192246.245 0.0 > Propers 1718.934378393635.973 0.1 > Impropers 134.502690 27976.560 0.0 > Pos. Restr.321.706434 16085.322 0.0 > Virial 424.7348267645.227 0.0 > Stop-CM 85.184882 851.849 0.0 > P-Coupling8484.719691 50908.318 0.0 > Calc-Ekin 848.794382 22917.448 0.0 > Lincs 313.720420 18823.225 0.0 > Lincs-Mat 1564.1465766256.586 0.0 > Constraint-V 8651.865815 69214.927 0.0 > Constraint-Vir 417.065668 10009.576 0.0 > Settle2674.808325 863963.089 0.2 > - > Total 487878233.910 100.0 > - > > > D O M A I N D E C O M P O S I T I O N S T A T I S T I C S > > av. #atoms communicated per step for force: 2 x 63413.7 > av. #atoms communicated per step for LINCS: 2 x 3922.5 > > Average load imbalance: 59.4 % > Part of the total run time spent waiting due to load imbalance: 5.0 % > > > R E A L C Y C L E A N D T I M E A C C O U N T I N G > > Computing: Nodes Th. Count Wall t (s) G-Cycles % > - > Domain decomp. 34 2500 42.792 1300.947 4.4 > DD comm. load 34 31 0.0000.014 0.0 > Neighbor search34 2501 33.076 1005.542 3.4 > Launch GPU ops.34 12 6.537 198.739 0.7 > Comm. coord. 34 47500 20.349 618.652 2.1 > Force 34 50001 75.093 2282.944 7.8 > Wait + Comm. F 34 50001 24.850 755.482 2.6 > PME mesh 34 50001 597.92518177.76062.0 > Wait GPU nonlocal 34 50001 9.862 299.813 1.0 > Wait GPU local 34 50001 0.2627.968 0.0 > NB X/F buffer ops. 34 195002 33.578 1020.833 3.5 > Write traj.34 12 0.506 15.385 0.1 > Update 34 50001 23.243
Re: [gmx-users] load imbalance in multiple GPU simulations
Hi Szilard, On Sun, Dec 8, 2013 at 2:48 PM, Szilárd Páll wrote: > Hi, > > That's unfortunate, but not unexpected. You are getting a 3x1x1 > decomposition where the "middle" cell has most of the protein, hence > most of the bonded forces to calculate, while the ones on the side > have little (or none). > > From which values can I tell this? > Currently, the only thing you can do is to try using more domains, > perhaps with manual decomposition (such that the initial domains will > contain as much protein as possible). This may not help much, though. > In extreme cases (e.g. small system), even using only two of the three > GPUs could improve performance Cheers, > -- > Szilárd > > > On Sun, Dec 8, 2013 at 8:10 PM, yunshi11 . wrote: > > Hi all, > > > > My conventional MD run (equilibration) of a protein in TIP3 water had the > > "Average load imbalance: 59.4 %" when running with 3 GPUs + 12 CPU cores. > > So I wonder how to tweak parameters to optimize the performance. > > > > End of the log file reads: > > > > .. > > M E G A - F L O P S A C C O U N T I N G > > > > NB=Group-cutoff nonbonded kernelsNxN=N-by-N cluster Verlet kernels > > RF=Reaction-Field VdW=Van der Waals QSTab=quadratic-spline table > > W3=SPC/TIP3p W4=TIP4p (single or pairs) > > V&F=Potential and force V=Potential only F=Force only > > > > Computing: M-Number M-Flops % > Flops > > > - > > Pair Search distance check 78483.330336706349.973 0.1 > > NxN QSTab Elec. + VdW [F] 11321254.234368 464171423.609 > 95.1 > > NxN QSTab Elec. + VdW [V&F] 114522.922048 6756852.401 > 1.4 > > 1,4 nonbonded interactions1645.932918148133.963 0.0 > > Calc Weights 25454.159073916349.727 0.2 > > Spread Q Bspline543022.060224 1086044.120 > 0.2 > > Gather F Bspline543022.060224 3258132.361 > 0.7 > > 3D-FFT 1138719.444112 9109755.553 > 1.9 > > Solve PME 353.129616 22600.295 0.0 > > Reset In Box 424.2275001272.682 > 0.0 > > CG-CoM 424.3971911273.192 > 0.0 > > Bonds 330.706614 19511.690 0.0 > > Angles1144.322886192246.245 0.0 > > Propers 1718.934378393635.973 0.1 > > Impropers 134.502690 27976.560 0.0 > > Pos. Restr.321.706434 16085.322 > 0.0 > > Virial 424.7348267645.227 > 0.0 > > Stop-CM 85.184882 851.849 > 0.0 > > P-Coupling8484.719691 50908.318 > 0.0 > > Calc-Ekin 848.794382 22917.448 0.0 > > Lincs 313.720420 18823.225 0.0 > > Lincs-Mat 1564.1465766256.586 > 0.0 > > Constraint-V 8651.865815 69214.927 0.0 > > Constraint-Vir 417.065668 10009.576 0.0 > > Settle2674.808325 863963.089 > 0.2 > > > - > > Total 487878233.910 > 100.0 > > > - > > > > > > D O M A I N D E C O M P O S I T I O N S T A T I S T I C S > > > > av. #atoms communicated per step for force: 2 x 63413.7 > > av. #atoms communicated per step for LINCS: 2 x 3922.5 > > > > Average load imbalance: 59.4 % > > Part of the total run time spent waiting due to load imbalance: 5.0 % > > > > > > R E A L C Y C L E A N D T I M E A C C O U N T I N G > > > > Computing: Nodes Th. Count Wall t (s) G-Cycles % > > > - > > Domain decomp. 34 2500 42.792 1300.947 > 4.4 > > DD comm. load 34 31 0.0000.014 > 0.0 > > Neighbor search34 2501 33.076 1005.542 > 3.4 > > Launch GPU ops.34 12 6.537 198.739 > 0.7 > > Comm. coord. 34 47500 20.349 618.652 > 2.1 > > Force 34 50001 75.093 2282.944 > 7.8 > > Wait + Comm. F 34 50001 24.850 755.482 > 2.6 > > PME mesh 34 50001 597.92518177.760 > 62.0 > > Wait GPU nonlocal 34 50001 9.862 299.813 > 1.0 > > Wait GPU local 34 5000
Re: [gmx-users] load imbalance in multiple GPU simulations
There is no value that tells you exactly that, but are clues. However, you can check in the log file the ratio of smallest and average (starting) cell size (the value also printed on the terminal with -v) and that will tell you how much did the DD shrink the middle cell. What you can also see is that if you run with -dd no, you'll get high load imbalance, but equal GPU (non-bonded load), but with -dd auto (or yes) you'll get much smaller load on the second GPU (use nvidia-smi). Cheers, -- Szilárd PS: You can somehow dump the PDB-s corresponding to the individual domains, but I don't exactly know how to do it (and that's rather low-level stuff anyway). On Mon, Dec 9, 2013 at 1:02 AM, yunshi11 . wrote: > Hi Szilard, > > > > > On Sun, Dec 8, 2013 at 2:48 PM, Szilárd Páll wrote: > >> Hi, >> >> That's unfortunate, but not unexpected. You are getting a 3x1x1 >> decomposition where the "middle" cell has most of the protein, hence >> most of the bonded forces to calculate, while the ones on the side >> have little (or none). >> >> From which values can I tell this? > > >> Currently, the only thing you can do is to try using more domains, >> perhaps with manual decomposition (such that the initial domains will >> contain as much protein as possible). This may not help much, though. >> In extreme cases (e.g. small system), even using only two of the three >> GPUs could improve performance > > Cheers, >> -- >> Szilárd >> >> >> On Sun, Dec 8, 2013 at 8:10 PM, yunshi11 . wrote: >> > Hi all, >> > >> > My conventional MD run (equilibration) of a protein in TIP3 water had the >> > "Average load imbalance: 59.4 %" when running with 3 GPUs + 12 CPU cores. >> > So I wonder how to tweak parameters to optimize the performance. >> > >> > End of the log file reads: >> > >> > .. >> > M E G A - F L O P S A C C O U N T I N G >> > >> > NB=Group-cutoff nonbonded kernelsNxN=N-by-N cluster Verlet kernels >> > RF=Reaction-Field VdW=Van der Waals QSTab=quadratic-spline table >> > W3=SPC/TIP3p W4=TIP4p (single or pairs) >> > V&F=Potential and force V=Potential only F=Force only >> > >> > Computing: M-Number M-Flops % >> Flops >> > >> - >> > Pair Search distance check 78483.330336706349.973 0.1 >> > NxN QSTab Elec. + VdW [F] 11321254.234368 464171423.609 >> 95.1 >> > NxN QSTab Elec. + VdW [V&F] 114522.922048 6756852.401 >> 1.4 >> > 1,4 nonbonded interactions1645.932918148133.963 0.0 >> > Calc Weights 25454.159073916349.727 0.2 >> > Spread Q Bspline543022.060224 1086044.120 >> 0.2 >> > Gather F Bspline543022.060224 3258132.361 >> 0.7 >> > 3D-FFT 1138719.444112 9109755.553 >> 1.9 >> > Solve PME 353.129616 22600.295 0.0 >> > Reset In Box 424.2275001272.682 >> 0.0 >> > CG-CoM 424.3971911273.192 >> 0.0 >> > Bonds 330.706614 19511.690 0.0 >> > Angles1144.322886192246.245 0.0 >> > Propers 1718.934378393635.973 0.1 >> > Impropers 134.502690 27976.560 0.0 >> > Pos. Restr.321.706434 16085.322 >> 0.0 >> > Virial 424.7348267645.227 >> 0.0 >> > Stop-CM 85.184882 851.849 >> 0.0 >> > P-Coupling8484.719691 50908.318 >> 0.0 >> > Calc-Ekin 848.794382 22917.448 0.0 >> > Lincs 313.720420 18823.225 0.0 >> > Lincs-Mat 1564.1465766256.586 >> 0.0 >> > Constraint-V 8651.865815 69214.927 0.0 >> > Constraint-Vir 417.065668 10009.576 0.0 >> > Settle2674.808325 863963.089 >> 0.2 >> > >> - >> > Total 487878233.910 >> 100.0 >> > >> - >> > >> > >> > D O M A I N D E C O M P O S I T I O N S T A T I S T I C S >> > >> > av. #atoms communicated per step for force: 2 x 63413.7 >> > av. #atoms communicated per step for LINCS: 2 x 3922.5 >> > >> > Average load imbalance: 59.4 % >> > Part of the total run time spent waiting due to load imbalance: 5.0 % >> > >> > >> > R E A L C Y C L E A N D T I M E A C C O U N T I N G >> > >> > Computing: Nodes Th. Count Wall t (s) G-Cycles %
[gmx-users] Compilation issue with F77_FUNC functions?
So, I'm trying to compile with MPI using mpich3. Previous installations worked, and installations without MPI worked. I'm getting errors like: /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: warning: parameter names (without types) in function declaration /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: error: function ‘F77_FUNC’ is initialized like a variable /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: warning: braces around scalar initializer /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: warning: (near initialization for ‘F77_FUNC’) /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: error: invalid initializer /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: error: (near initialization for ‘F77_FUNC’) /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: warning: excess elements in scalar initializer /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: warning: (near initialization for ‘F77_FUNC’) /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: warning: excess elements in scalar initializer /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: warning: (near initialization for ‘F77_FUNC’) And it keeps on like that for a long while. Any suggestions? Perhaps something wrong in the way mpicc is handling Fortran code? -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Compilation issue with F77_FUNC functions?
Hi, what compiler is used by mpicc? What does "mpicc -showme" and "mpicc --version" show? Does it help to uncomment the line containing F77_FUNC int src/config.h.cmakein? Roland On Sun, Dec 8, 2013 at 7:44 PM, Michael Shirts wrote: > So, I'm trying to compile with MPI using mpich3. Previous > installations worked, and installations without MPI worked. I'm > getting errors like: > > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: > warning: parameter names (without types) in function declaration > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: > error: function ‘F77_FUNC’ is initialized like a variable > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: > warning: braces around scalar initializer > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: > warning: (near initialization for ‘F77_FUNC’) > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > error: invalid initializer > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > error: (near initialization for ‘F77_FUNC’) > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > warning: excess elements in scalar initializer > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > warning: (near initialization for ‘F77_FUNC’) > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > warning: excess elements in scalar initializer > /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: > warning: (near initialization for ‘F77_FUNC’) > > And it keeps on like that for a long while. > > Any suggestions? Perhaps something wrong in the way mpicc is handling > Fortran code? > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or > send a mail to gmx-users-requ...@gromacs.org. > > > > > -- ORNL/UT Center for Molecular Biophysics cmb.ornl.gov 865-241-1537, ORNL PO BOX 2008 MS6309 -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Compilation issue with F77_FUNC functions?
Apologies! I certainly didn't post enough information. mpicc is using: gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-3) Which is exactly what the non-mpi versions used, which did not have this problem. I ran cmake with: cmake ../gromacs -DGMX_MPI=ON -DGMX_GPU=OFF -DGMX_DOUBLE=ON -DGMX_CPU_ACCELERATION=AVX_256 -DCMAKE_INSTALL_PREFIX=/h3/n1/shirtsgroup/gromacs_46/install -DFFTW_INCLUDE_DIR=/h3/n1/shirtsgroup/software/fft3w/include -DFFTW_LIBRARY='/h3/n1/shirtsgroup/software/fft3w/lib/libfftw3.a;/h3/n1/shirtsgroup/software/fft3w/lib/libfftw3.so;/usr/lib64/libm.so;/share/apps/mpich3/gnu/lib/libmpich.so;' All the extra libraries for FFTWF appear to be necessary for some reason, but I don't think that's it . . . Commentinf out the line F77_FUNC int src/config.h.cmakein and rerunning cmake did not change anything. On Sun, Dec 8, 2013 at 10:29 PM, Roland Schulz wrote: > Hi, > > what compiler is used by mpicc? What does "mpicc -showme" and "mpicc > --version" show? Does it help to uncomment the line containing F77_FUNC int > src/config.h.cmakein? > > Roland > > > On Sun, Dec 8, 2013 at 7:44 PM, Michael Shirts wrote: > >> So, I'm trying to compile with MPI using mpich3. Previous >> installations worked, and installations without MPI worked. I'm >> getting errors like: >> >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: >> warning: parameter names (without types) in function declaration >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: >> error: function ‘F77_FUNC’ is initialized like a variable >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: >> warning: braces around scalar initializer >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: >> warning: (near initialization for ‘F77_FUNC’) >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> error: invalid initializer >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> error: (near initialization for ‘F77_FUNC’) >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> warning: excess elements in scalar initializer >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> warning: (near initialization for ‘F77_FUNC’) >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> warning: excess elements in scalar initializer >> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >> warning: (near initialization for ‘F77_FUNC’) >> >> And it keeps on like that for a long while. >> >> Any suggestions? Perhaps something wrong in the way mpicc is handling >> Fortran code? >> -- >> 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.gromacs.org/Support/Mailing_Lists >> >> * For (un)subscribe requests visit >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or >> send a mail to gmx-users-requ...@gromacs.org. >> >> >> >> >> > > > -- > ORNL/UT Center for Molecular Biophysics cmb.ornl.gov > 865-241-1537, ORNL PO BOX 2008 MS6309 > -- > 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.gromacs.org/Support/Mailing_Lists > > * For (un)subscribe requests visit > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a > mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Compilation issue with F77_FUNC functions?
And FWIW, it's being compiled on CentOS 6.4. On Sun, Dec 8, 2013 at 11:16 PM, Michael Shirts wrote: > Apologies! I certainly didn't post enough information. mpicc is using: > > gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-3) > > Which is exactly what the non-mpi versions used, which did not have > this problem. > > I ran cmake with: > > cmake ../gromacs -DGMX_MPI=ON -DGMX_GPU=OFF -DGMX_DOUBLE=ON > -DGMX_CPU_ACCELERATION=AVX_256 > -DCMAKE_INSTALL_PREFIX=/h3/n1/shirtsgroup/gromacs_46/install > -DFFTW_INCLUDE_DIR=/h3/n1/shirtsgroup/software/fft3w/include > -DFFTW_LIBRARY='/h3/n1/shirtsgroup/software/fft3w/lib/libfftw3.a;/h3/n1/shirtsgroup/software/fft3w/lib/libfftw3.so;/usr/lib64/libm.so;/share/apps/mpich3/gnu/lib/libmpich.so;' > > All the extra libraries for FFTWF appear to be necessary for some > reason, but I don't think that's it . . . > > Commentinf out the line F77_FUNC int src/config.h.cmakein and > rerunning cmake did not change anything. > > On Sun, Dec 8, 2013 at 10:29 PM, Roland Schulz wrote: >> Hi, >> >> what compiler is used by mpicc? What does "mpicc -showme" and "mpicc >> --version" show? Does it help to uncomment the line containing F77_FUNC int >> src/config.h.cmakein? >> >> Roland >> >> >> On Sun, Dec 8, 2013 at 7:44 PM, Michael Shirts wrote: >> >>> So, I'm trying to compile with MPI using mpich3. Previous >>> installations worked, and installations without MPI worked. I'm >>> getting errors like: >>> >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: >>> warning: parameter names (without types) in function declaration >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:53: >>> error: function ‘F77_FUNC’ is initialized like a variable >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: >>> warning: braces around scalar initializer >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:56: >>> warning: (near initialization for ‘F77_FUNC’) >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> error: invalid initializer >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> error: (near initialization for ‘F77_FUNC’) >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> warning: excess elements in scalar initializer >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> warning: (near initialization for ‘F77_FUNC’) >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> warning: excess elements in scalar initializer >>> /h3/n1/shirtsgroup/gromacs_46/gromacs/src/gmxlib/cinvsqrtdata.c:57: >>> warning: (near initialization for ‘F77_FUNC’) >>> >>> And it keeps on like that for a long while. >>> >>> Any suggestions? Perhaps something wrong in the way mpicc is handling >>> Fortran code? >>> -- >>> 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.gromacs.org/Support/Mailing_Lists >>> >>> * For (un)subscribe requests visit >>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or >>> send a mail to gmx-users-requ...@gromacs.org. >>> >>> >>> >>> >>> >> >> >> -- >> ORNL/UT Center for Molecular Biophysics cmb.ornl.gov >> 865-241-1537, ORNL PO BOX 2008 MS6309 >> -- >> 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.gromacs.org/Support/Mailing_Lists >> >> * For (un)subscribe requests visit >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a >> mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] Fwd: Normal Mode Analysis
Dear users, I am trying to do NMA First i did energy minimization using .mdp file with conjugate gradient method, Next i calculated hessian matrix by using integrator = nm Then i calculated the eigen vectors from 7 to 100 using g_nmeig to analyze eigen vectors i use the commands g_anaeig_d -s nm.tpr -f em-c.gro -v eigenvec.trr -eig eigenval.xvg -proj proj-ev1.xvg -extr ev1.pdb -rmsf rmsf-ev1.xvg -first 7 -last 7 -nframes 30 by visuvalizing ev1.pdb, i did not found any motion in the protein. what is the mistake i have done? Is it correct procedure for doing NM analysis? How to analyze the eigen values obtained from hessian matrix? regards M.SathishKumar -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] Fw: the crashed run
thanks for your reply Have a Great Day On Monday, December 9, 2013 1:35 AM, Justin Lemkul wrote: On 12/8/13 6:59 AM, Mahboobeh Eslami wrote: > thanks for your reply > please suggest the best command for restart a crashed run. > i use following command > mdrun -v -deffnm md -cpi md > is this command good or not? If the desired .cpt file is md.cpt, then mdrun -deffnm md -cpi suffices. If you need a previous .cpt file (i.e. md_prev.cpt), then you must specify its name explicitly. Depending on why the run crashed, simply restarting may not be worthwhile. If the crash was due to the system blowing up, it is a waste of time and you should investigate the crash. If it was a hardware failure, disk error, etc. then continuing from a checkpoint is fine. -Justin > thanks > > > > > On Sunday, December 8, 2013 2:18 PM, "jkrie...@mrc-lmb.cam.ac.uk" > wrote: > > _prev.cpt might be ok. I think it writes those in case the current cpt causes > problems > > > On 8 Dec 2013, at 09:57, Mahboobeh Eslami wrote: > >> hi GMX users >> i use > gromacs 4.6.3 double precision for protein ligand complex during the 20 ns. my run crashed, I'm not > sure that restart my run from cpt file or run a new production. >> is the result of started again run a reliable like a non crashed run. >> >> thanks for your help >> -- >> 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.gromacs.org/Support/Mailing_Lists >> >> * For (un)subscribe requests visit >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a >> mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] charge correction in topology file from PRODRG
dear justin thanks for your help On Monday, December 9, 2013 1:44 AM, Justin Lemkul wrote: On 12/8/13 6:18 AM, XAvier Periole wrote: > > For gromos ff you can also use the ATB server. From Alan Mark on Brisbane. It > combine typography and non-bonded parameters all at once. > > It is not perfect but pretty good. > For those wondering about some potential implications of topology errors and how to start going about fixing them, as well as an overview of some common QM calculations one can do to try to calculate charges for new groups: http://pubs.acs.org/doi/abs/10.1021/ci100335w -Justin >> On Dec 8, 2013, at 8:27, Mahboobeh Eslami wrote: >> >> hi all my friends >> I use PRODRG and antechmber for building topology and coordinate files for >> my ligand separately. >> i want to use GROMOS force field so i must to use the topology from PRODRG >> server. can i use the topology of antechamber for charge correction in >> topology fiel from PRODRG . >> In general, are special principlesessential for charge correction in >> topology file from PRODRG >> thanks for your help >> -- >> 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.gromacs.org/Support/Mailing_Lists >> >> * For (un)subscribe requests visit >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a >> mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
Re: [gmx-users] charge correction in topology file from PRODRG
dear justin thanks for your help On Monday, December 9, 2013 1:44 AM, Justin Lemkul wrote: On 12/8/13 6:18 AM, XAvier Periole wrote: > > For gromos ff you can also use the ATB server. From Alan Mark on Brisbane. It > combine typography and non-bonded parameters all at once. > > It is not perfect but pretty good. > For those wondering about some potential implications of topology errors and how to start going about fixing them, as well as an overview of some common QM calculations one can do to try to calculate charges for new groups: http://pubs.acs.org/doi/abs/10.1021/ci100335w -Justin >> On Dec 8, 2013, at 8:27, Mahboobeh Eslami wrote: >> >> hi all my friends >> I use PRODRG and antechmber for building topology and coordinate files for >> my ligand separately. >> i want to use GROMOS force field so i must to use the topology from PRODRG >> server. can i use the topology of antechamber for charge correction in >> topology fiel from PRODRG . >> In general, are special principlesessential for charge correction in >> topology file from PRODRG >> thanks for your help >> -- >> 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.gromacs.org/Support/Mailing_Lists >> >> * For (un)subscribe requests visit >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a >> mail to gmx-users-requ...@gromacs.org. -- == Justin A. Lemkul, Ph.D. Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 601 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 == -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org. -- 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.gromacs.org/Support/Mailing_Lists * For (un)subscribe requests visit https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-requ...@gromacs.org.
[gmx-users] simulation using tabulated dihedral potentials.
I am facing some weird problem while using tabulated potentials (esp. dihedral tabulated potentials) in gromacs 4.6.3. I am trying to generate a coarse grained (CG) trajectory using the tabulated potentials generated for the atomistic simulations using VOTCA. When I donot use the dihedral tabulated potentials, the simulations seems to proceed smoothly. The problem arises when the dihedral potentials are incorporated. I have added a link of all the files generated during the undermentioned runs (https://www.dropbox.com/s/yumzdufuys1ifdr/with-dih.tar) . It also includes the input files. Here, I list the problem and how do they occur : 1. $mdrun_463 -v -nice 0 -cpt 1 -cpi state.cpt -nt 2 -pin on (1st Run) >& ver1.txt Progress smoothly for few steps (step 536400). Shows the following error : A list of missing interactions: Tab. Dih. of720 missing 1 Molecule type 'POLCAR' the first 10 missing interactions, except for exclusions: Tab. Dih. atoms6789 global 858 859 860 861 --- Program mdrun_463, VERSION 4.6.3 Source code file: /tmp/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393 Fatal error: 1 of the 2400 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.5 nm) or the two-body cut-off distance (1.5 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck For more information and tips for troubleshooting, please check the GROMACS website at http://www.gromacs.org/Documentation/Errors --- "Sometimes Life is Obscene" (Black Crowes 2. Then I continue the run using the state.cpt file $mdrun_463 -v -nice 0 -cpt 1 -cpi state.cpt -nt 2 -pin on >& ver2.txt Further progressed for few more steps (step 973100). Following error is produced here: A list of missing interactions: Tab. Angles of800 missing 1 Tab. Dih. of720 missing 1 Molecule type 'POLCAR' the first 10 missing interactions, except for exclusions: Tab. Dih. atoms9 10 11 12 global 237 238 239 240 Tab. Angles atoms 10 11 12 global 238 239 240 Back Off! I just backed up dd_dump_err_0_n1.pdb to ./#dd_dump_err_0_n1.pdb.1# Back Off! I just backed up dd_dump_err_0_n0.pdb to ./#dd_dump_err_0_n0.pdb.1# --- Program mdrun_463, VERSION 4.6.3 Source code file: /tmp/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393 Fatal error: 2 of the 2400 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.5 nm) or the two-body cut-off distance (1.5 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck For more information and tips for troubleshooting, please check the GROMACS website at http://www.gromacs.org/Documentation/Errors --- "She's a Good Sheila Bruce" (Monty Python) 3. Again continued the simulation $mdrun_463 -v -nice 0 -cpt 1 -cpi state.cpt -nt 2 -pin on >& ver3.txt Simulation proceeds for further for step 3692700. The again crashed with the following output: A list of missing interactions: Tab. Angles of800 missing 1 Tab. Dih. of720 missing 2 Molecule type 'POLCAR' the first 10 missing interactions, except for exclusions: Tab. Dih. atoms6789 global 306 307 308 309 Tab. Angles atoms789 global 307 308 309 Tab. Dih. atoms789 10 global 307 308 309 310 Back Off! I just backed up dd_dump_err_0_n0.pdb to ./#dd_dump_err_0_n0.pdb.2# Back Off! I just backed up dd_dump_err_0_n1.pdb to ./#dd_dump_err_0_n1.pdb.2# --- Program mdrun_463, VERSION 4.6.3 Source code file: /tmp/gromacs-4.6.3/src/mdlib/domdec_top.c, line: 393 Fatal error: 3 of the 2400 bonded interactions could not be calculated because some atoms involved moved further apart than the multi-body cut-off distance (1.5 nm) or the two-body cut-off distance (1.5 nm), see option -rdd, for pairs and tabulated bonds also see option -ddcheck For more information and tips for troubleshooting, please check the GROMACS website at http://www.gromacs.org/Documentation/Errors --- "Though the path of the comet is sure, it's constitution is not" (Peter Hammill) In this way the simulations proceed. At the end it finally stops at step 71768380 and no further continuation is possible. Error at this stage : Started mdrun on node 0 Mon Dec 9 12:24:52 2013 Step Time Lambda 71768380 358841.90.0 Energies (kJ/mol) Tab. BondsTab. Angles Tab. Dih.