https://community.intel.com/t5/Intel-MPI-Library/Running-coupled-executables-with-different-thread-counts-using/m-p/1138802#M5857 <P>&nbsp;</P> <P>Under LSF how can I run mutiple executables with different thread counts and still use the nodes efficiently?</P> <P>Currently I have to do</P> <P>#BSUB -R [ptile=7]</P> <P>#BSUB -R affinity[core(4)]</P> <P>mpirun&nbsp; -n 8 -env OMP_NUM_THREADS=2 ./hellope : -n 12 -env OMP_NUM_THREADS=1 ./hellope : -n 8 -env OMP_NUM_THREADS=4 ./hellope</P> <P>This will yield a otal number of 60 threads.&nbsp; For a node with 28 processors it would take 3 nodes, but since I have one executable with 4 threads I have to tile it to allow for the maximum thread count so as to not overlap processes on cores.&nbsp; This means I need to use another node even thougn I don't need all of the cores.&nbsp; Is there a way I can pack this onto a node? or place the tasks where I want them?&nbsp; I thought maybe something with using I_MPI_JOB_RESPECT_PROCESS_PLACEMENT and maybe defining the affinity somewhere else besides the job card?&nbsp; Any thoughts would be appreciated.</P> <P>&nbsp;</P> Wed, 03 Oct 2018 20:38:22 GMT Pasti__Carolyn 2018-10-03T20:38:22Z https://community.intel.com/t5/Intel-MPI-Library/Running-coupled-executables-with-different-thread-counts-using/m-p/1138802#M5857 <P>&nbsp;</P> <P>Under LSF how can I run mutiple executables with different thread counts and still use the nodes efficiently?</P> <P>Currently I have to do</P> <P>#BSUB -R [ptile=7]</P> <P>#BSUB -R affinity[core(4)]</P> <P>mpirun&nbsp; -n 8 -env OMP_NUM_THREADS=2 ./hellope : -n 12 -env OMP_NUM_THREADS=1 ./hellope : -n 8 -env OMP_NUM_THREADS=4 ./hellope</P> <P>This will yield a otal number of 60 threads.&nbsp; For a node with 28 processors it would take 3 nodes, but since I have one executable with 4 threads I have to tile it to allow for the maximum thread count so as to not overlap processes on cores.&nbsp; This means I need to use another node even thougn I don't need all of the cores.&nbsp; Is there a way I can pack this onto a node? or place the tasks where I want them?&nbsp; I thought maybe something with using I_MPI_JOB_RESPECT_PROCESS_PLACEMENT and maybe defining the affinity somewhere else besides the job card?&nbsp; Any thoughts would be appreciated.</P> <P>&nbsp;</P> Wed, 03 Oct 2018 20:38:22 GMT https://community.intel.com/t5/Intel-MPI-Library/Running-coupled-executables-with-different-thread-counts-using/m-p/1138802#M5857 Pasti__Carolyn 2018-10-03T20:38:22Z