Re: [gmx-users] Increasing performance of siumulation in cluster
There are lots of things you might do, but when we don't know what's in the simulation or anything about the hardware, nobody can tell. Mark On Thu, May 30, 2013 at 2:16 PM, Sainitin Donakonda saigr...@gmail.comwrote: Hi all, I recently ran 20ns simulation in linux cluster. Used following script for MD simulation #This is the first simulation MD.mdp file contains 20 ns setup grompp -f MD.mdp -c npt.gro -t npt.cpt -p topol.top -n index.ndx -o MD_first10.tpr mpirun -n 16 mdrun -s MD_first10.tpr -deffnm MD_first10 -np 16 #This extends 10 ns simulation tpbconv -s MD_first10.tpr -o md_extended.tpr -extend 1 mpirun -n 16 mdrun -s md_extended.tpr -deffnm MD_first10 -cpi MD_first10.cpt -append -np 16 and bsub file #!/bin/bash #BSUB -J testgromacs # the job's name/array job #BSUB -W 120:00 # max. wall clock time in hh:mm #BSUB -n 16,16 # number of processors Min,Max #BSUB -o /home/sample/output_%J.log # output file #BSUB -e /home/sample/errors_%J.log # error file #BSUB -M 8192 #Memory limit in MB echo Started at `date` echo cd /home/sample/ echo Running gromacs test in `pwd` ./MD.sh echo Finished at `date` * * * * *It ran for total 5 days but didnt give all 20ns simulation ..So i checked log file for performance * *Parallel run - timing based on wallclock.* * * * NODE (s) Real (s) (%)* * Time: 359019.178 359019.178100.0* * 4d03h43:39* * (Mnbf/s) (GFlops) (ns/day) (hour/ns)* *Performance:445.015 24.708 2.407 9.973* * * * * *It looks like this gave very performance 2.4 ns per 9 hours..this looks very low for me.. Can any body tell me how to increase performance of simulation* Thanks Nitin -- 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] Increasing performance of siumulation in cluster
Hi Mark, I forgot to mention about simulation it is protein ligand simulation which contains one protein and ligand with 2 solute molecules..and regarding hardware of cluster which is in my university as follows CPUs AMD Opteron 6274 number of cores 5888 th. peak performance 51.8 TFlops compute nodes 4-way nodes *Saxonid* with 64 cores nodes with 64 GB RAM 48 nodes with 128 GB RAM 24 nodes with 256 GB RAM 12 nodes with 512 GB RAM 8 Can you now suggest some methods which i can use to optimize to increase performance Thanks, Nitin On Fri, May 31, 2013 at 10:54 AM, Mark Abraham mark.j.abra...@gmail.comwrote: There are lots of things you might do, but when we don't know what's in the simulation or anything about the hardware, nobody can tell. Mark On Thu, May 30, 2013 at 2:16 PM, Sainitin Donakonda saigr...@gmail.com wrote: Hi all, I recently ran 20ns simulation in linux cluster. Used following script for MD simulation #This is the first simulation MD.mdp file contains 20 ns setup grompp -f MD.mdp -c npt.gro -t npt.cpt -p topol.top -n index.ndx -o MD_first10.tpr mpirun -n 16 mdrun -s MD_first10.tpr -deffnm MD_first10 -np 16 #This extends 10 ns simulation tpbconv -s MD_first10.tpr -o md_extended.tpr -extend 1 mpirun -n 16 mdrun -s md_extended.tpr -deffnm MD_first10 -cpi MD_first10.cpt -append -np 16 and bsub file #!/bin/bash #BSUB -J testgromacs # the job's name/array job #BSUB -W 120:00 # max. wall clock time in hh:mm #BSUB -n 16,16 # number of processors Min,Max #BSUB -o /home/sample/output_%J.log # output file #BSUB -e /home/sample/errors_%J.log # error file #BSUB -M 8192 #Memory limit in MB echo Started at `date` echo cd /home/sample/ echo Running gromacs test in `pwd` ./MD.sh echo Finished at `date` * * * * *It ran for total 5 days but didnt give all 20ns simulation ..So i checked log file for performance * *Parallel run - timing based on wallclock.* * * * NODE (s) Real (s) (%)* * Time: 359019.178 359019.178100.0* * 4d03h43:39* * (Mnbf/s) (GFlops) (ns/day) (hour/ns)* *Performance:445.015 24.708 2.407 9.973* * * * * *It looks like this gave very performance 2.4 ns per 9 hours..this looks very low for me.. Can any body tell me how to increase performance of simulation* Thanks Nitin -- 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
Re: [gmx-users] Increasing performance of siumulation in cluster
On Fri, May 31, 2013 at 11:30 AM, Sainitin Donakonda saigr...@gmail.comwrote: Hi Mark, I forgot to mention about simulation it is protein ligand simulation which contains one protein and ligand with 2 solute molecules.. So assuming you mean solvent, about 60K atoms. and regarding hardware of cluster which is in my university as follows CPUs AMD Opteron 6274 number of cores 5888 th. peak performance 51.8 TFlops compute nodes 4-way nodes *Saxonid* with 64 cores nodes with 64 GB RAM 48 nodes with 128 GB RAM 24 nodes with 256 GB RAM 12 nodes with 512 GB RAM 8 Links are better than copy-and-paste... https://doc.zih.tu-dresden.de/hpc-wiki/bin/view/Compendium/HardwareAtlas Can you now suggest some methods which i can use to optimize to increase performance Your hardware is quite useful. You should be getting better performance. Exactly how to do that depends on the simulation (e.g. PME or not, group or Verlet kernels). You should start with http://www.gromacs.org/Documentation/Acceleration_and_parallelization and experiment with the balance of MPI processes and OpenMP threads per node of your hardware. Look at the timing breakdown at the end of the .log file, and pay attention to all the information grompp and mdrun writes! Mark -- 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