Re: [gmx-users] GTX 960 vs Tesla K40
Thanks for the heads up. With the K40c instead of GTX 960 here's what I did and here are the results: 1. Enabled persistence mode and overclocked the card via nvidia-smi: http://acceleware.com/blog/gpu-boost-nvidias-tesla-k40-gpus 2. Offloaded PME's FFT to GPU (which wasn't the case with GTX 960), this brough the "pme mesh / force" ratio to something like 1.07. The result is a solid increase in performance on a small-ish system (20K atoms): 90 ns/day instead of 65-70. I don't use this box for anything except prototyping, but still the swap + tweaks were pretty useful. Alex On 6/15/2018 1:20 PM, Szilárd Páll wrote: Hi, Regarding the K40 vs GTX 960 question, the K40 will likely be a bit faster (though it'l consume more power if that matters). The difference will be at most 20% in total performance, I think -- and with small systems likely negligible (as a smaller card with higher clocks is more efficient at small tasks than a large card with lower clocks). Regarding the load balance note, you are correct, the "pme mesh/force" means the ratio of time spent in computing PME forces on a separate task/rank and the rest of the forces (including nonbonded, bonded, etc.). With GPU offload this is a bit more tricky as the observed time is the time spent waiting for the GPU results, but the take-away is the same: when a run shows "pme mesh/force" far from 1, there is imbalance affecting performance. However, note that with a single GPU I've yet to see a case where you get better performance by running multiple ranks rather than simply running OpenMP-only. Also note that what a "weak GPU" can case-by-case, so I recommend taking the 1-2 minutes to do a short run and check for a certain hardware + simulation setup is it better to offload all of PME or keep the FFTs on the CPU. We'll do our best to automate more of these choices, but for now if you care about performance it's useful to test before doing long runs. Cheers, -- Szilárd On Thu, Jun 14, 2018 at 2:09 AM, Alex wrote: Question: in the DD output (md.log) that looks like "DD step xx pme mesh/force 1.229," what is the ratio? Does it mean the pme calculations take longer by the shown factor than the nonbonded interactions? With GTX 960, the ratio is consistently ~0.85, with Tesla K40 it's ~1.25. My mdrun line contains -pmefft cpu (per Szilard's advice for weak GPUs, I believe). Would it then make sense to offload the fft to the K40? Thank you, Alex On Wed, Jun 13, 2018 at 4:53 PM, Alex wrote: So, swap, then? Thank you! On Wed, Jun 13, 2018 at 4:49 PM, paul buscemi wrote: flops trumps clock speed….. On Jun 13, 2018, at 3:45 PM, Alex wrote: Hi all, I have an old "prototyping" box with a 4-core Xeon and an old GTX 960. We have a Tesla K40 laying around and there's only one PCIE slot available in this machine. Would it make sense to swap the cards, or is it already bottlenecked by the CPU? I compared the specs and 960 has a higher clock speed, while K40's FP performance is better. Should I swap the GPUs? Thanks, Alex -- 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. -- 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] gromacs.org_gmx-users Digest, Vol 170, Issue 57
Thank you Shreyas, Pedro and Mark Abraham for forwarding my initial message to the program's author who responded to me Stéphane -- -- Message: 3 Date: Sat, 16 Jun 2018 19:45:28 +0200 From: Shreyas Kaptan To: gmx-us...@gromacs.org Subject: Re: [gmx-users] About the "g_contacts" program for gromacs Message-ID: Content-Type: text/plain; charset="UTF-8" In the paper it mentions that: This paper and its associated computer program are available via the Computer Physics Communication homepage on ScienceDirect ( http://www.sciencedirect.com/science/journal/00104655). Hope that helps. On Thu, Jun 14, 2018 at 5:53 PM ABEL Stephane wrote: > Hi there, > > Do you know where I can find the program g_contacts developed by Christian > Blau & Helmut Grubmuller and described in the following paper "g_contacts: > Fast contact search in bio-molecular ensemble data. Computer Physics > Communications > Volume 184, Issue 12, December 2013, Pages 2856-2859" > https://www.sciencedirect.com/science/article/pii/S0010465513002464. I > would like to test it. In addition, does this tool work with the 5.1.X or > newer versions of gromacs. (It was initially developed for gromacs 4.6) . > > Thanks > > A bient?t > > St?phane > > -- > 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. > -- Shreyas Sanjay Kaptan -- Message: 4 Date: Sat, 16 Jun 2018 19:04:39 +0100 From: Pedro Deira To: gmx-us...@gromacs.org Subject: Re: [gmx-users] About the "g_contacts" program for gromacs Message-ID: Content-Type: text/plain; charset="UTF-8" I never managed to get it working with 5.x, I have a separate 4.x installation to run g_contacts. p. On Thu, 14 Jun 2018, 16:53 ABEL Stephane, wrote: > Hi there, > > Do you know where I can find the program g_contacts developed by Christian > Blau & Helmut Grubmuller and described in the following paper "g_contacts: > Fast contact search in bio-molecular ensemble data. Computer Physics > Communications > Volume 184, Issue 12, December 2013, Pages 2856-2859" > https://www.sciencedirect.com/science/article/pii/S0010465513002464. I > would like to test it. In addition, does this tool work with the 5.1.X or > newer versions of gromacs. (It was initially developed for gromacs 4.6) . > > Thanks > > A bient?t > > St?phane > > -- > 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. End of gromacs.org_gmx-users Digest, Vol 170, Issue 57 ** -- 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] membrane-protein system by using charmm36 ff
Hi. Maybe you already know this but you can also build the whole embedded system with charmm-gui. Also, your parameters appear reasonable to me at first glance. As for the equilibration, that is a system specific question. If you have a "simple" uniform lipid content in the bilayer I would say from my experience, that the equilibration depends on the lipid heads and tails. Large heads and long tails generally imply a longer equilibration. Mixed lipids can require up to "microseconds" worth of equilibratio. I would take the saturation to a nearly fixed value of the Area per lipid and the bilayer thickness as an indication that it is safe to consider the "equilibration" enough. Do not use the 0.495 ns as some timescale. It is in fact quite short. On Sun, Jun 17, 2018 at 1:25 PM Olga Press wrote: > Dear Gromacs users, > I'm new in the field of Molecular Dynamics especially in using Gromacs. > I have several questions regarding mdp file and I'll be very grateful if > you can help me with them. > I'm using a membrane-protein system with Charmm36 ff. After I have > constructed bilayer membrane by using CHARMM-GUI membrane builder I have > run the README file as it, without changing the equilibration time (total > equilibration time of 0.475ns). Followed by embedded protein into the > membrane by using g_membed and performed solvation and minimization of the > entire system as was described in the KALP15-DPPC tutorial by Dr. Justin > A.Lemkul. > > those are my questions: > 1. Does the pre-equilibration of 0.475ns is enough before embedding protein > into the membrane and followed by long equilibration of the whole system > for 200ns by using NVT followed by NPT equilibration? > > 2. I've read that when using CHARMM36 ff in gromacs is better to switch the > following parameters > constraints = h-bonds > cutoff-scheme = Verlet > vdwtype = cutoff > vdw-modifier = force-switch > rlist = 1.2 > rvdw = 1.2 > rvdw-switch = 1.0 > coulombtype = PME > rcoulomb = 1.2 > DispCorr = no > > I'm using the original mdout.mdp files produces by gromacs.Are those > parameters optimal for a membrane-protein system or just for the lipids? > > Thank you all for your help. > Olga > -- > 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. > -- Shreyas Sanjay Kaptan -- 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] membrane-protein system by using charmm36 ff
Dear Gromacs users, I'm new in the field of Molecular Dynamics especially in using Gromacs. I have several questions regarding mdp file and I'll be very grateful if you can help me with them. I'm using a membrane-protein system with Charmm36 ff. After I have constructed bilayer membrane by using CHARMM-GUI membrane builder I have run the README file as it, without changing the equilibration time (total equilibration time of 0.475ns). Followed by embedded protein into the membrane by using g_membed and performed solvation and minimization of the entire system as was described in the KALP15-DPPC tutorial by Dr. Justin A.Lemkul. those are my questions: 1. Does the pre-equilibration of 0.475ns is enough before embedding protein into the membrane and followed by long equilibration of the whole system for 200ns by using NVT followed by NPT equilibration? 2. I've read that when using CHARMM36 ff in gromacs is better to switch the following parameters constraints = h-bonds cutoff-scheme = Verlet vdwtype = cutoff vdw-modifier = force-switch rlist = 1.2 rvdw = 1.2 rvdw-switch = 1.0 coulombtype = PME rcoulomb = 1.2 DispCorr = no I'm using the original mdout.mdp files produces by gromacs.Are those parameters optimal for a membrane-protein system or just for the lipids? Thank you all for your help. Olga -- 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.
