On Fri, Apr 2, 2021, 20:58 Satyen Dhamankar <[email protected]> wrote:
> Christoph, > > I also ran some ibi tests on the urea-water tutorial example. > I am running into the same issues. The IBI is running at an incredibly > slow pace. From what I have gathered, it seems to be a problem when you > have more than one bead type in the coarse-grained system that IBI gets > incredibly slow. I have heard that tabulated potentials are incredibly slow > in GMX... Could that be a contributing factor? > Tabulated interactions are slow but not more than a factor 2. More bead types means more energy groups and that could be the issue, but for multiple tables you need multiple energy groups. You could try an older gromacs version... But not many other users have reported that issue before, so I would investigate more why gromacs is slow first. Christoph > > On Friday, April 2, 2021 at 3:10:49 PM UTC-4 Satyen Dhamankar wrote: > >> Christoph, >> >> Thank you for the advice. >> I have added -nt 12 in the following manner: >> >> <inverse> >> <!-- 200*0.00831451 gromacs units --> >> <kBT>1.66290</kBT> >> <!-- use gromacs as simulation program --> >> <program>gromacs</program> >> <!-- gromacs specific options --> >> <gromacs> >> <!-- trash so many frames at the beginning --> >> <equi_time>20</equi_time> >> <!-- grid for table*.xvg !--> >> <table_bins>0.002</table_bins> >> <!-- cut the potential at this value (gromacs bug) --> >> <pot_max>1000000</pot_max> >> <!-- extend the tables to this value --> >> <table_end>2.5</table_end> >> <mdrun.opts>-nt 12</mdrun.opts> >> </gromacs> >> >> however, i am getting the same message: >> "Running on 1 node with total 28 cores, 28 logical cores" >> and also runs on them >> "Initializing Domain Decomposition on 28 ranks" >> >> I have attached my settings.xml file to this document (the only change i >> have made is the -nt 12 part). >> >> >> Speaking of, I made the change and wrote -nt 8 in the methanol tutorial, >> and that one worked fine... >> On Friday, April 2, 2021 at 10:20:55 AM UTC-4 Christoph Junghans wrote: >> >>> On Fri, Apr 2, 2021 at 7:16 AM Satyen Dhamankar <[email protected]> >>> wrote: >>> > >>> > The gromacs run. It is strange because the spce ibi did not take as >>> long. >>> > I have attached my md.log file. After 2.5 hours, only 11999 steps were >>> completed for a 3000 molecule simulation... >>> Hmm, that might be more a question for the gromacs user mailing list. >>> >>> From your md.log, I see that Gromacs detected 28 cores >>> "Running on 1 node with total 28 cores, 28 logical cores" >>> and also runs on them >>> "Initializing Domain Decomposition on 28 ranks" >>> >>> However you only requested 12 cores in your batch script: >>> #SBATCH --ntasks-per-node=12 >>> >>> If the other cores are used by another user that could lead to >>> performance problems as gromacs always runs as slow as the slowest >>> core. >>> gromacs has an option to limit the number cores, I believe "-nt 12", >>> which you can hand over to gromacs from the xml file using >>> cg.inverse.gromacs.mdrun.opts. >>> >>> Christoph >>> >>> > >>> > On Friday, April 2, 2021 at 8:47:18 AM UTC-4 Christoph Junghans wrote: >>> >> >>> >> On Thu, Apr 1, 2021 at 21:57 Satyen Dhamankar <[email protected]> >>> wrote: >>> >>> >>> >>> Hello, >>> >>> >>> >>> I have been running the propane IBI tutorial for propane. The IBI >>> for propane takes quite some time to run. Step_001 has taken more than an >>> hour to complete. >>> >>> >>> >>> This is the slurm script I have used to run this job: >>> >>> >>> >>> #!/bin/bash >>> >>> # >>> >>> #SBATCH --job-name=R1.5-CG # create a short name for your job >>> >>> #SBATCH --qos=vshort # _quality of service_ >>> >>> #SBATCH --nodes=1 # node count >>> >>> #SBATCH --ntasks-per-node=12 # number of tasks per node >>> >>> #SBATCH --cpus-per-task=1 # cpu-cores per task (>1 if multi-threaded >>> tasks) >>> >>> #SBATCH --mem=10GB # total memory requested >>> >>> ##SBATCH --mem-per-cpu=4G # memory per cpu-core (4G per cpu-core is >>> default) >>> >>> #SBATCH --gres=gpu:1 # number of gpurs per node >>> >>> #SBATCH --time=2:30:00 # total run time limit (HH:MM:SS) >>> >>> #SBATCH --mail-type=all # send email on job start, end, and fail >>> >>> module purge >>> >>> module load intel/19.0/64/19.0.5.281 # for running gromacs >>> >>> module load rh/devtoolset/7 >>> >>> module load intel-mkl/2019.5/5/64 >>> >>> module load cmake/3.x >>> >>> # run the job >>> >>> set -e >>> >>> # keep track of the last executed command >>> >>> trap 'last_command=$current_command; current_command=$BASH_COMMAND' >>> DEBUG >>> >>> # echo an error message before exiting >>> >>> #trap 'echo "\"${last_command\" command filed with exit code $?." ' >>> EXIT >>> >>> bash run.sh >>> >>> >>> >>> Is taking such a long time for a sample IBI run normal? >>> >> >>> >> Where does it get stuck? In the gromacs run or after that? >>> >> >>> >> Christoph >>> >>> >>> >>> -- >>> >>> Join us on Slack: https://join.slack.com/t/votca/signup >>> >>> --- >>> >>> You received this message because you are subscribed to the Google >>> Groups "votca" group. >>> >>> To unsubscribe from this group and stop receiving emails from it, >>> send an email to [email protected]. >>> >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/votca/dadf2e09-20dd-4e63-8eaf-f748a3d390a3n%40googlegroups.com. >>> >>> >> >>> >> -- >>> >> Christoph Junghans >>> >> Web: http://www.compphys.de >>> > >>> > -- >>> > Join us on Slack: https://join.slack.com/t/votca/signup >>> > --- >>> > You received this message because you are subscribed to the Google >>> Groups "votca" group. >>> > To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> > To view this discussion on the web visit >>> https://groups.google.com/d/msgid/votca/0847a33c-636b-4a09-82c0-4fdd5a63c1e5n%40googlegroups.com. >>> >>> >>> >>> >>> -- >>> Christoph Junghans >>> Web: http://www.compphys.de >>> >> -- > Join us on Slack: https://join.slack.com/t/votca/signup > --- > You received this message because you are subscribed to the Google Groups > "votca" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/votca/75b980dd-748f-4cfd-b0fc-b9a5ccb0741cn%40googlegroups.com > <https://groups.google.com/d/msgid/votca/75b980dd-748f-4cfd-b0fc-b9a5ccb0741cn%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- Join us on Slack: https://join.slack.com/t/votca/signup --- You received this message because you are subscribed to the Google Groups "votca" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/votca/CAHG27e44m5tT-JzeLZusuSLb-Hb2BgxLzxPVJ2usQybNXWs%3DSw%40mail.gmail.com.
