Re: [gmx-users] Performance of 4.6.1 vs. 4.5.5
On Sat, Mar 9, 2013 at 6:53 AM, Christopher Neale chris.ne...@mail.utoronto.ca wrote: Dear users: I am seeing a 140% performance boost when moving from gromacs 4.5.5 to 4.6.1 when I run a simulation on a single node. However, I am only seeing a 110% performance boost when running on multiple nodes. Does anyone else see this? Note that I am not using the verlet cutoff scheme. What's the processor and network for those runs? I'm not sure that this is a problem, but I was surprised to see how big the difference was between 1 and 2 nodes, while for 2-10 nodes I saw a reliable 10% performance boost. Not sure what you mean by reliable 10% performance boost. Reporting actual ns/day rates would be clearer. Is a 140% performance boost a factor of 1.4 more ns/day or a factor of 2.4 more ns/day? Please note that, while I compiled the fftw (with sse2) and gromacs 4.6.1, I did not compile the 4.5.5 version that I am comparing to (or its fftw) so the difference might be in compilation options. Indeed. Still, I wonder why the benefits of 4.6.1 are so fantastic on 1 node but fall off to good-but-not-amazing on 1 node. Finding the answer would start by examining the changes in the timing breakdowns in your .log files. Switching from using in-memory MPI to network MPI is a significant cost on busy/weak networks. The system is about 43K atoms. I have not tested this with other systems or cutoffs. My mdp file follows. Thank you for any advice. Your system is probably not calculating energies very much. 4.6 uses force-only kernels if that's all you need from it. Mark Chris. constraints = all-bonds lincs-iter = 1 lincs-order = 6 constraint_algorithm = lincs integrator = sd dt = 0.002 tinit = 0 nsteps = 25 nstcomm = 1 nstxout = 25 nstvout = 25 nstfout = 25 nstxtcout = 5 nstenergy = 5 nstlist = 10 nstlog=0 ns_type = grid vdwtype = switch rlist = 1.0 rlistlong = 1.6 rvdw = 1.5 rvdw-switch = 1.4 rcoulomb = 1.0 coulombtype = PME ewald-rtol = 1e-5 optimize_fft = yes fourierspacing = 0.12 fourier_nx = 0 fourier_ny = 0 fourier_nz = 0 pme_order = 4 tc_grps = System tau_t = 1.0 ld_seed = -1 ref_t = 310 gen_temp = 310 gen_vel = yes unconstrained_start = no gen_seed = -1 Pcoupl = berendsen pcoupltype = semiisotropic tau_p = 4 4 compressibility = 4.5e-5 4.5e-5 ref_p = 1.0 1.0 dispcorr = EnerPres -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Gromacs with Intel Xeon Phi coprocessors ?
No idea. It was not a development target for 4.6, and there are no explicit plans for considering Xeon Phi at this time. It would be interesting to hear whether people can get benefit from them from our existing OpenMP parallelism support for (particularly) PME + Verlet kernels. Roughly speaking, 4.6 delivered a lot of performance improvements within (and in spite of) an ageing C code base. The plans for 5.0 target a very limited number of new features, chief of which is a transition to using C++. We are targeting the kind of massive utilisation of threads that seems to us likely to feature in the future of HPC. So far, using Intel's TBB as a test bed is something we are considering, though we may end up rolling our own code instead. Either way, a straightforward port of GROMACS that works well on Xeon Phi seems fairly likely to me. Just not this year ;-) (Unless anyone has hardware and developer time to donate!) Mark On Sat, Mar 9, 2013 at 2:43 AM, Christopher Neale chris.ne...@mail.utoronto.ca wrote: Dear users: does anybody have any experience with gromacs on a cluster in which each node is composed of 1 or 2 x86 processors plus an Intel Xeon Phi coprocessor? Can gromacs make use of the xeon phi coprocessor? If not, does anybody know if that is in the pipeline? Thank you, Chris. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Performance of 4.6.1 vs. 4.5.5
As Mark said, we need concrete details to answer the question: - log files (all four of them: 1/2 nodes, 4.5/4.6) - hardware (CPUs, network) - compilers The 4.6 log files contain much of the second and third point except the network. Note that you can compare the performance summary table's entries one by one and see what has changed. I suspect that the answer is simply load imbalance, but we'll have to see the numbers to know. -- Szilárd On Sat, Mar 9, 2013 at 3:00 PM, Mark Abraham mark.j.abra...@gmail.comwrote: On Sat, Mar 9, 2013 at 6:53 AM, Christopher Neale chris.ne...@mail.utoronto.ca wrote: Dear users: I am seeing a 140% performance boost when moving from gromacs 4.5.5 to 4.6.1 when I run a simulation on a single node. However, I am only seeing a 110% performance boost when running on multiple nodes. Does anyone else see this? Note that I am not using the verlet cutoff scheme. What's the processor and network for those runs? I'm not sure that this is a problem, but I was surprised to see how big the difference was between 1 and 2 nodes, while for 2-10 nodes I saw a reliable 10% performance boost. Not sure what you mean by reliable 10% performance boost. Reporting actual ns/day rates would be clearer. Is a 140% performance boost a factor of 1.4 more ns/day or a factor of 2.4 more ns/day? Please note that, while I compiled the fftw (with sse2) and gromacs 4.6.1, I did not compile the 4.5.5 version that I am comparing to (or its fftw) so the difference might be in compilation options. Indeed. Still, I wonder why the benefits of 4.6.1 are so fantastic on 1 node but fall off to good-but-not-amazing on 1 node. Finding the answer would start by examining the changes in the timing breakdowns in your .log files. Switching from using in-memory MPI to network MPI is a significant cost on busy/weak networks. The system is about 43K atoms. I have not tested this with other systems or cutoffs. My mdp file follows. Thank you for any advice. Your system is probably not calculating energies very much. 4.6 uses force-only kernels if that's all you need from it. Mark Chris. constraints = all-bonds lincs-iter = 1 lincs-order = 6 constraint_algorithm = lincs integrator = sd dt = 0.002 tinit = 0 nsteps = 25 nstcomm = 1 nstxout = 25 nstvout = 25 nstfout = 25 nstxtcout = 5 nstenergy = 5 nstlist = 10 nstlog=0 ns_type = grid vdwtype = switch rlist = 1.0 rlistlong = 1.6 rvdw = 1.5 rvdw-switch = 1.4 rcoulomb = 1.0 coulombtype = PME ewald-rtol = 1e-5 optimize_fft = yes fourierspacing = 0.12 fourier_nx = 0 fourier_ny = 0 fourier_nz = 0 pme_order = 4 tc_grps = System tau_t = 1.0 ld_seed = -1 ref_t = 310 gen_temp = 310 gen_vel = yes unconstrained_start = no gen_seed = -1 Pcoupl = berendsen pcoupltype = semiisotropic tau_p = 4 4 compressibility = 4.5e-5 4.5e-5 ref_p = 1.0 1.0 dispcorr = EnerPres -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] query regarding mk_angndx
-- Forwarded message -- From: Kavyashree M hmkv...@gmail.com Date: Fri, Mar 8, 2013 at 10:45 PM Subject: query regarding mk_angndx To: Discussion list for GROMACS users gmx-users@gromacs.org Dear users, I used mkang_ndx to create an index file with dihedral angles. Input was: mk_angndx -s a.tpr -n angle.ndx -type dihedral output angle.ndx read like this - [ Phi=180.0_2_43.93 ] 52018192237353627323031 395957586176747566716970 ... According to my understanding the numbers indicate the 4 atoms defining the particular dihedral angle. But when I checked the pdb file for these atoms- ATOM 5 CA MET A 1 111.430 40.170 113.130 1.00 0.00 ATOM 18 C MET A 1 112.060 41.020 112.030 1.00 0.00 ATOM 19 O MET A 1 111.910 42.240 112.010 1.00 0.00 ATOM 20 N GLN A 2 112.940 40.430 111.220 1.00 0.00 I could not make out how this defines phi? Kindly clarify my confusion. Thank you kavya -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] Re: Calculation of coordination number for lithium ion
Please keep all Gromacs-related questions on the gmx-user mailing list. I am not a private tutor. On 3/8/13 7:20 PM, qzg00...@nifty.com wrote: Dear Dr. Justin Using g_rdf , I want to calculate coordination number of water molecule around Li+ in a Li salt aqueous solution. In order to calculate the radial distribution function, two Groups should be selected as 'Reference' and '1 group'. For the Li+ , which groups should I select among 0(system) -9(sideChain-H)? I think that I should select System for Reference and again System for 1 group. Is this correct? A system-system RDF is likely useless. If you want the RDF of water around lithium, it should be easy to select lithium as the reference group and water (or even better, OW atoms in an index group of your creation) as the group around that central group. Selecting protein-related terms (i.e. within the first 10 default groups) makes no sense in light of what you're trying to measure. -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
Re: [gmx-users] Re: RDF for INVACCUO simulation
On 3/9/13 12:34 AM, Keerthana S.P Periasamy wrote: DEAR ALL For RDF calculation in invaccuo simulation I have given the command as follows ./g_rdf_d -f outputdsvaccum_md.trr -s outputdsvaccum_md.tpr -n index.ndx -nopbc -o rdfss.xvg I am getting the graph which I have attached with this mail. Can you suggest whether my command and the graph is correct and I am proceeding in the right way. Attachments are not allowed to the list, but you can provide links to where files can be downloaded. That said, no one is going to be able to tell you anything about the sensibility of your data from such a (nonexistent) description. Nor is anyone likely to invest time in doing your homework for you ;) -Justin -- Justin A. Lemkul, Ph.D. Research Scientist Department of Biochemistry Virginia Tech Blacksburg, VA jalemkul[at]vt.edu | (540) 231-9080 http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] error in gromacs 4.0.7-Source code file: domdec.c, line: 5888
Dear users I used gromacs for my system include CNT-water-ion-protein (400,000atom), I used grompp without error. after used mdrun with 16 node on cluster, I get this error: Reading file nvt.tpr, VERSION 4.0.7 (single precision) Loaded with Money NOTE: Periodic molecules: can not easily determine the required minimum bonded cut-off, using half the non-bonded cut-off Will use 15 particle-particle and 1 PME only nodes This is a guess, check the performance at the end of the log file --- Program mdrun, VERSION 4.0.7 Source code file: domdec.c, line: 5888 Fatal error: There is no domain decomposition for 15 nodes that is compatible with the given box and a minimum cell size of 3.75 nm Change the number of nodes or mdrun option -rdd or -dds Look in the log file for details on the domain decomposition --- Good Music Saves your Soul (Lemmy) Error on node 0, will try to stop all the nodes Halting parallel program mdrun on CPU 0 out of 16 --- my box size is 14*14*18. I used mdrun -rdd 1 , gromacs run without error is it correct or not? I get this error in NVT (Em run without error. Best Regards. Hamid Mosaddeghi -- View this message in context: http://gromacs.5086.n6.nabble.com/error-in-gromacs-4-0-7-Source-code-file-domdec-c-line-5888-tp5006245.html Sent from the GROMACS Users Forum mailing list archive at Nabble.com. -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
[gmx-users] Re: gromacs VERSION 4.0.7-There is no domain decomposition.....
Dear Hamid, please post these kind of questions on the users and not the developers list. 2013/3/9 Hamid Mosaddeghi hamid5920...@yahoo.com: Dear users I used gromacs for my system include CNT-water-ion-protein (400,000atom), I used grompp without error. after used mdrun with 16 node on cluster, I get this error: Reading file nvt.tpr, VERSION 4.0.7 (single precision) Loaded with Money NOTE: Periodic molecules: can not easily determine the required minimum bonded cut-off, using half the non-bonded cut-off Will use 15 particle-particle and 1 PME only nodes This is a guess, check the performance at the end of the log file --- Program mdrun, VERSION 4.0.7 Source code file: domdec.c, line: 5888 Fatal error: There is no domain decomposition for 15 nodes that is compatible with the given box and a minimum cell size of 3.75 nm Change the number of nodes or mdrun option -rdd or -dds Look in the log file for details on the domain decomposition --- Good Music Saves your Soul (Lemmy) Error on node 0, will try to stop all the nodes Halting parallel program mdrun on CPU 0 out of 16 --- my box size is 14*14*18. I used mdrun -rdd 1 , gromacs run without error is it correct or not? If you are sure that the maximum distance of your bonded interactions is 1nm, you can do that. It seems like mdrun was not able to find a decomposition automatically, try to give one by hand: $ mdrun -dd 2 2 4 Btw, your gromacs is 2 major versions behind, it might be a good idea to update. Christoph I get this error in NVT (Em run without error). Best Regards. Hamid Mosaddeghi -- View this message in context: http://gromacs.5086.n6.nabble.com/gromacs-VERSION-4-0-7-There-is-no-domain-decomposition-tp5006246.html Sent from the GROMACS Developers Forum mailing list archive at Nabble.com. -- gmx-developers mailing list gmx-develop...@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-developers Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-developers-requ...@gromacs.org. -- Christoph Junghans Web: http://www.compphys.de -- gmx-users mailing listgmx-users@gromacs.org http://lists.gromacs.org/mailman/listinfo/gmx-users * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting! * Please don't post (un)subscribe requests to the list. Use the www interface or send it to gmx-users-requ...@gromacs.org. * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
