job scheduling - Cray XE6 parallel tasks / parallel tasks per node -

in job submission script (cray xe6) difference between 'number of parallel tasks' (-n) , 'number of parallel tasks per node' (-n)?

im inexperienced in googling gave:

it looks mpi parameter in example:

# launch parallel job using aprun. # run executable my_mpi_executable.x using total # of 2048 parallel tasks, 32 tasks assigned per node. aprun -n 2048 -n 32 ./my_mpi_executable.x arg1 arg2 

from cray brochure:

each cray xe6 blade includes 4 compute nodes high scalability in small footprint – 128 processor cores per blade or 3,072 processor cores per cabinet. each compute node has 2 amd opteron™ 6300 series processors (16 cores) coupled own memory , gemini communication interface , designed efficiently run 32 mpi tasks. alternately, can programmed run openmp within compute node , mpi between nodes.

from info above, n can 3072 full cabinet , n can 32 2 opterons(6300 series).