Re: [gmx-users] Vacuum simulation in Gromacs 5
Hi Szilárd, thank you for you reaction! I prepared a simple input which reproduces the behavior, see https://owncloud.cesnet.cz/public.php?service=files&t=9e8fc3924fd109f6d48a552e20ceaf11 With -nt 1 it successfully finishes in ca 12 seconds, with -nt 2 it gets stuck after SETTLE citation. Regards, Jan 2015-06-15 21:51 GMT+02:00 Szilárd Páll : > On Sun, Jun 14, 2015 at 6:54 PM, Jan Jirsák wrote: >> Hi, >> >> I did the test and found out that -nt 8 is even slower than -nt 1 !! > > FYI: "-nt" is mostly backward compatibility option and for clarity > it's best to use either -ntmpi or -ntomp, depending on what you mean. > >> However, I think that simulation hasn't even properly started with 8 >> threads and got stuck somewhere in the beginning. >> >> Details: >> I used a short run (1000 steps) for testing. mdrun -nt 1 finished >> after ca 11 hours, whereas not a single checkpoint or run log record >> have been saved for mdrun -nt 8 (and simulation is still running). Not >> even starting energies were displayed - last record in a log file is a >> SETTLE citation. > > That's probably a stuck which should not happen and I'd consider that > a bug. Can you share your input files? > >> In top I noticed both processes used up to 100% CPU, but considerably >> differed in the use of memory. Relevant lines: >> PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND >> 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 >> mdrun >> 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun > > That definitely does not look good. > > BTW, in contrast with Mark's opinion I do think top is indicative. If > you wanted to use 8 cores and told mdrun to do so but you get load > only on one core, you can be (almost) certain that something is not > right. > -- 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] Vacuum simulation in Gromacs 5
On Sun, Jun 14, 2015 at 6:54 PM, Jan Jirsák wrote: > Hi, > > I did the test and found out that -nt 8 is even slower than -nt 1 !! FYI: "-nt" is mostly backward compatibility option and for clarity it's best to use either -ntmpi or -ntomp, depending on what you mean. > However, I think that simulation hasn't even properly started with 8 > threads and got stuck somewhere in the beginning. > > Details: > I used a short run (1000 steps) for testing. mdrun -nt 1 finished > after ca 11 hours, whereas not a single checkpoint or run log record > have been saved for mdrun -nt 8 (and simulation is still running). Not > even starting energies were displayed - last record in a log file is a > SETTLE citation. That's probably a stuck which should not happen and I'd consider that a bug. Can you share your input files? > In top I noticed both processes used up to 100% CPU, but considerably > differed in the use of memory. Relevant lines: > PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND > 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 > mdrun > 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun That definitely does not look good. BTW, in contrast with Mark's opinion I do think top is indicative. If you wanted to use 8 cores and told mdrun to do so but you get load only on one core, you can be (almost) certain that something is not right. -- Szilárd > The first listed process (117710) is the single thread simulation (-nt 1). > > Thank you for any insight, > Jan > > 2015-06-12 14:35 GMT+02:00 Mark Abraham : >> Hi, >> >> Top need not be indicative. Run the same with -nt 1 and observe whether the >> performance changes. And don't run anything else on these cores. >> >> Mark > -- > 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] Vacuum simulation in Gromacs 5
Ok, thanks -- I see I will have to go back to version 4.5 or 4.6 where my code worked flawlessly (particle decomposition is available there). Regarding "flying away" -- sometimes I NEED my system to disintegrate to fragments which fly away;) Momentum is of course conserved, cause they fly in opposite directions. Thank you for the links, I will have a look. With kind regards Jan 2015-06-14 19:32 GMT+02:00 V.V.Chaban : > There are just two options for such simulations. Either one uses a > code supporting particle decomposition or adds a vacuum layer. > > The fragments will not fly away provided that momentum is conserved. > > Some comments on why and what to select may be present here: > > http://pubs.acs.org/doi/abs/10.1021/jz500563q > > http://pubs.acs.org/doi/abs/10.1021/jz300405q > > http://pubs.acs.org/doi/abs/10.1021/jz201190j > > > > > > > On Sun, Jun 14, 2015 at 2:24 PM, Jan Jirsák wrote: >> Thank you for a reply - I could try that when everything else fails, but >> (i) my system can disintegrate and its fragments can fly really long >> distances apart, so the layer of vacuum would have to be considerably >> (and unpredictably) thick - the same applies to cutoffs, >> (ii) in principle I don't see why one should bother with periodic >> images in a simulation which is inherently non-periodic. >> Regards, Jan >> >> >> 2015-06-14 19:01 GMT+02:00 V.V.Chaban : >>> Why don't you simply surround your molecule with a layer of vacuum and >>> use all periodic set of details as normally? This is what people do in >>> the plane-wave codes when the so-called 'cluster representation' is >>> desired. >>> >>> >>> >>> >>> On Sun, Jun 14, 2015 at 1:54 PM, Jan Jirsák wrote: Hi, I did the test and found out that -nt 8 is even slower than -nt 1 !! However, I think that simulation hasn't even properly started with 8 threads and got stuck somewhere in the beginning. Details: I used a short run (1000 steps) for testing. mdrun -nt 1 finished after ca 11 hours, whereas not a single checkpoint or run log record have been saved for mdrun -nt 8 (and simulation is still running). Not even starting energies were displayed - last record in a log file is a SETTLE citation. In top I noticed both processes used up to 100% CPU, but considerably differed in the use of memory. Relevant lines: PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 mdrun 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun The first listed process (117710) is the single thread simulation (-nt 1). Thank you for any insight, Jan 2015-06-12 14:35 GMT+02:00 Mark Abraham : > Hi, > > Top need not be indicative. Run the same with -nt 1 and observe whether > the > performance changes. And don't run anything else on these cores. > > Mark -- 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] Vacuum simulation in Gromacs 5
There are just two options for such simulations. Either one uses a code supporting particle decomposition or adds a vacuum layer. The fragments will not fly away provided that momentum is conserved. Some comments on why and what to select may be present here: http://pubs.acs.org/doi/abs/10.1021/jz500563q http://pubs.acs.org/doi/abs/10.1021/jz300405q http://pubs.acs.org/doi/abs/10.1021/jz201190j On Sun, Jun 14, 2015 at 2:24 PM, Jan Jirsák wrote: > Thank you for a reply - I could try that when everything else fails, but > (i) my system can disintegrate and its fragments can fly really long > distances apart, so the layer of vacuum would have to be considerably > (and unpredictably) thick - the same applies to cutoffs, > (ii) in principle I don't see why one should bother with periodic > images in a simulation which is inherently non-periodic. > Regards, Jan > > > 2015-06-14 19:01 GMT+02:00 V.V.Chaban : >> Why don't you simply surround your molecule with a layer of vacuum and >> use all periodic set of details as normally? This is what people do in >> the plane-wave codes when the so-called 'cluster representation' is >> desired. >> >> >> >> >> On Sun, Jun 14, 2015 at 1:54 PM, Jan Jirsák wrote: >>> Hi, >>> >>> I did the test and found out that -nt 8 is even slower than -nt 1 !! >>> However, I think that simulation hasn't even properly started with 8 >>> threads and got stuck somewhere in the beginning. >>> >>> Details: >>> I used a short run (1000 steps) for testing. mdrun -nt 1 finished >>> after ca 11 hours, whereas not a single checkpoint or run log record >>> have been saved for mdrun -nt 8 (and simulation is still running). Not >>> even starting energies were displayed - last record in a log file is a >>> SETTLE citation. >>> >>> In top I noticed both processes used up to 100% CPU, but considerably >>> differed in the use of memory. Relevant lines: >>> PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND >>> 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 >>> mdrun >>> 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun >>> >>> The first listed process (117710) is the single thread simulation (-nt 1). >>> >>> Thank you for any insight, >>> Jan >>> >>> 2015-06-12 14:35 GMT+02:00 Mark Abraham : Hi, Top need not be indicative. Run the same with -nt 1 and observe whether the performance changes. And don't run anything else on these cores. Mark >>> -- >>> 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] Vacuum simulation in Gromacs 5
Thank you for a reply - I could try that when everything else fails, but (i) my system can disintegrate and its fragments can fly really long distances apart, so the layer of vacuum would have to be considerably (and unpredictably) thick - the same applies to cutoffs, (ii) in principle I don't see why one should bother with periodic images in a simulation which is inherently non-periodic. Regards, Jan 2015-06-14 19:01 GMT+02:00 V.V.Chaban : > Why don't you simply surround your molecule with a layer of vacuum and > use all periodic set of details as normally? This is what people do in > the plane-wave codes when the so-called 'cluster representation' is > desired. > > > > > On Sun, Jun 14, 2015 at 1:54 PM, Jan Jirsák wrote: >> Hi, >> >> I did the test and found out that -nt 8 is even slower than -nt 1 !! >> However, I think that simulation hasn't even properly started with 8 >> threads and got stuck somewhere in the beginning. >> >> Details: >> I used a short run (1000 steps) for testing. mdrun -nt 1 finished >> after ca 11 hours, whereas not a single checkpoint or run log record >> have been saved for mdrun -nt 8 (and simulation is still running). Not >> even starting energies were displayed - last record in a log file is a >> SETTLE citation. >> >> In top I noticed both processes used up to 100% CPU, but considerably >> differed in the use of memory. Relevant lines: >> PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND >> 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 >> mdrun >> 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun >> >> The first listed process (117710) is the single thread simulation (-nt 1). >> >> Thank you for any insight, >> Jan >> >> 2015-06-12 14:35 GMT+02:00 Mark Abraham : >>> Hi, >>> >>> Top need not be indicative. Run the same with -nt 1 and observe whether the >>> performance changes. And don't run anything else on these cores. >>> >>> Mark >> -- >> 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] Vacuum simulation in Gromacs 5
Why don't you simply surround your molecule with a layer of vacuum and use all periodic set of details as normally? This is what people do in the plane-wave codes when the so-called 'cluster representation' is desired. On Sun, Jun 14, 2015 at 1:54 PM, Jan Jirsák wrote: > Hi, > > I did the test and found out that -nt 8 is even slower than -nt 1 !! > However, I think that simulation hasn't even properly started with 8 > threads and got stuck somewhere in the beginning. > > Details: > I used a short run (1000 steps) for testing. mdrun -nt 1 finished > after ca 11 hours, whereas not a single checkpoint or run log record > have been saved for mdrun -nt 8 (and simulation is still running). Not > even starting energies were displayed - last record in a log file is a > SETTLE citation. > > In top I noticed both processes used up to 100% CPU, but considerably > differed in the use of memory. Relevant lines: > PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND > 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 > mdrun > 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun > > The first listed process (117710) is the single thread simulation (-nt 1). > > Thank you for any insight, > Jan > > 2015-06-12 14:35 GMT+02:00 Mark Abraham : >> Hi, >> >> Top need not be indicative. Run the same with -nt 1 and observe whether the >> performance changes. And don't run anything else on these cores. >> >> Mark > -- > 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] Vacuum simulation in Gromacs 5
Hi, I did the test and found out that -nt 8 is even slower than -nt 1 !! However, I think that simulation hasn't even properly started with 8 threads and got stuck somewhere in the beginning. Details: I used a short run (1000 steps) for testing. mdrun -nt 1 finished after ca 11 hours, whereas not a single checkpoint or run log record have been saved for mdrun -nt 8 (and simulation is still running). Not even starting energies were displayed - last record in a log file is a SETTLE citation. In top I noticed both processes used up to 100% CPU, but considerably differed in the use of memory. Relevant lines: PID USER PR NIVIRTRESSHR S %CPU %MEM TIME+ COMMAND 117710 jan 20 0 5313776 4.642g 9940 R 98.9 3.7 52:47.19 mdrun 117686 jan 20 0 547344 24456 8712 R 92.7 0.0 54:48.92 mdrun The first listed process (117710) is the single thread simulation (-nt 1). Thank you for any insight, Jan 2015-06-12 14:35 GMT+02:00 Mark Abraham : > Hi, > > Top need not be indicative. Run the same with -nt 1 and observe whether the > performance changes. And don't run anything else on these cores. > > Mark -- 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] Vacuum simulation in Gromacs 5
Hi, Top need not be indicative. Run the same with -nt 1 and observe whether the performance changes. And don't run anything else on these cores. Mark On Fri, 12 Jun 2015 14:20 Jan Jirsák wrote: > Hi, > > I have one more problem with running this system with thread-MPI > (tested in both 5.0.4 and 5.0.5 on two different machines). When I set > everything as you advised me, it runs, however top command shows only > 100% load - i.e., only single CPU is used (and it is really very very > slow), despite log files confirm the desired number of MPI threads was > used. > > Thanks, Jan > -- > 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] Vacuum simulation in Gromacs 5
Hi, I have one more problem with running this system with thread-MPI (tested in both 5.0.4 and 5.0.5 on two different machines). When I set everything as you advised me, it runs, however top command shows only 100% load - i.e., only single CPU is used (and it is really very very slow), despite log files confirm the desired number of MPI threads was used. Thanks, Jan -- 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] Vacuum simulation in Gromacs 5
Hi, That particular optimization hasn't been implemented in the multi-domain case. It's on the table for post 5.1, however. Mark On Thu, 11 Jun 2015 17:45 Jan Jirsák wrote: > David van der Spoel writes: > > Use grid search in any case. It supports vacuum. > > Thank you very much, it seems to work. I wonder, does nstlist variable have > any relevance in this case? I mean, here all particles interact with one > another, so it should be sufficient to build neighbor list just once, but > still Gromacs does not allow me to set nstlist=0. > > Thanks, Jan > > > > -- > 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] Vacuum simulation in Gromacs 5
Justin Lemkul writes: > > Use mdrun -nt 1 > Thank you for a quick reply - however, I really need to parallelize - single CPU run would take ages;) Regards, Jan -- 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] Vacuum simulation in Gromacs 5
David van der Spoel writes: > Use grid search in any case. It supports vacuum. Thank you very much, it seems to work. I wonder, does nstlist variable have any relevance in this case? I mean, here all particles interact with one another, so it should be sufficient to build neighbor list just once, but still Gromacs does not allow me to set nstlist=0. Thanks, Jan -- 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] Vacuum simulation in Gromacs 5
On 6/11/15 5:53 AM, Jan Jirsák wrote: Hello everyone, what is the correct setup for simulations with no PBC and no cuttoffs in Gromacs 5.0.4? In versions 4.5 and 4.6 i used nstlist = 0 ns_type = simple pbc = no This no longer works, as I get the error: "Domain decomposition does not support simple neighbor searching, use grid searching or run with one MPI rank" (... and particle decomposition i no longer available.) However, when I run mdrun -nt 16 -ntmpi 1 (is this correct specification for 1 MPI rank?) I get the error that OpenMP can be used only with cutoffscheme=Verlet, which in turn is not available for pbc = no. Grid searching is nonsense in this situation, as cuttoffs are infinite. I seem to be in a dead end. Does anybody know a solution? Use mdrun -nt 1 -Justin -- == Justin A. Lemkul, Ph.D. Ruth L. Kirschstein NRSA Postdoctoral Fellow Department of Pharmaceutical Sciences School of Pharmacy Health Sciences Facility II, Room 629 University of Maryland, Baltimore 20 Penn St. Baltimore, MD 21201 jalem...@outerbanks.umaryland.edu | (410) 706-7441 http://mackerell.umaryland.edu/~jalemkul == -- 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] Vacuum simulation in Gromacs 5
On 11/06/15 11:53, Jan Jirsák wrote: Hello everyone, what is the correct setup for simulations with no PBC and no cuttoffs in Gromacs 5.0.4? In versions 4.5 and 4.6 i used nstlist = 0 ns_type = simple pbc = no This no longer works, as I get the error: "Domain decomposition does not support simple neighbor searching, use grid searching or run with one MPI rank" (... and particle decomposition i no longer available.) However, when I run mdrun -nt 16 -ntmpi 1 (is this correct specification for 1 MPI rank?) I get the error that OpenMP can be used only with cutoffscheme=Verlet, which in turn is not available for pbc = no. Grid searching is nonsense in this situation, as cuttoffs are infinite. I seem to be in a dead end. Does anybody know a solution? Use grid search in any case. It supports vacuum. Thank you, Jan -- 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.
