search

lanl / benchmarks

Benchmarks
BSD 3-Clause "New" or "Revised" License
14 stars 6 forks source link

Microbenchmark proofread suggestions #68

Closed JDTruj2018 closed 10 months ago

JDTruj2018 commented 10 months ago

66

See the comments in #66

JDTruj2018 commented 10 months ago

STREAM questions:

Which optimization flags were used? I assumed -O2 -fopenmp as is done here: https://github.com/jeffhammond/STREAM/blob/master/Makefile Were MPI ranks bound to cores? I assumed yes, and the srun commands now include --cpu-bind=core Was the stream-mpi.exe executable used? I assumed yes, and the srun commands now reflect that Was the STREAM_ARRAY_SIZE set to 40,000,000 rather than 40? That appears to give similar results to what is on the page (I used gcc 9.4.0 and openmpi/3.1.6 on Darwin's spr-hbm partition) STREAM changes:

Defined compile time pre-processor constants (-DSTREAM_ARRAY_SIZE=40000000 -DNTIMES=20) in the build section along with optimization flags (-O2 -fopenmp) Added example make command to the build section Added export OMP_NUM_THREADS=1 to the running section commands Added --cpu-bind=core to the srun command Changed the executable in the running examples from ./stream to ./stream-mpi.exe Fixed typo where STREAM_ARRAY_SIZE=40 is now STREAM_ARRAY_SIZE=40000000

dmageeLANL commented 10 months ago

This is great. Thank you for making the changes!