Intel® oneAPI HPC Toolkit
Get help with building, analyzing, optimizing, and scaling high-performance computing (HPC) applications.
Announcements
This community is designed for sharing of public information. Please do not share Intel or third-party confidential information here.
1921 Discussions

Profiling MPI applilcation with Vtune

seongyun_k_
Beginner
164 Views

Hi, folks

I'd like to profile my MPI application with Vtune.

In ordered to see the inter-node behaviors,I definitely need to use '-gtool' options to aggregate the profiled result into one file.

1) When I run the application without profiling, the following command works perfect:

  • $ mpiexec.hydra -genvall -n 8 -machinefile /home/my_name/machines ARGS1 ARGS2 ...

2) The following command also does the job (running multiple MPI processes on a machine). I can see the aggregated results of them.

  • $ mpiexec.hydra -genvall -gtool "amplxe-cl -c hotspots -r /prf_log_dir/1:all=node-wide" -n 8 ARGS1 ARGS2 ...

3) However, when I add '-machinefile (~)' option to specify the nodes to run, the profiled result data are not combined together. Each of the result file shows profiled result for each MPI process run on that node. 

  • $ mpiexec.hydra -genvall -n 8 -machinefile /home/my_name/machines -gtool "amplxe-cl -c hotspots -r /dir_to_result/result1:all=node-wide" ARGS1 ARGS2 ...

Am I doing something wrong here? I googled a lot and read lots of documents but I cannot get any glimpse on what's going on here...

 

 

0 Kudos
0 Replies
Reply