Difference between revisions of "MPAS-Ocean"

From Tau Wiki
Jump to: navigation, search
Line 10: Line 10:
 
The MPAS-Ocean code has been modified to use TAU as the timers, rather than the internal timers. This provides for both MPI performance measurement as well as PAPI counters.
 
The MPAS-Ocean code has been modified to use TAU as the timers, rather than the internal timers. This provides for both MPI performance measurement as well as PAPI counters.
  
The MPAS-Ocean developers have collected profiles on Hopper, with 192 to 16800 processes, using MPI only (no OpenMP yet).
+
The MPAS-Ocean developers have collected profiles on Hopper, with 192 to 16800 processes, using MPI only (no OpenMP yet). In addition, full callpath and communication matrix profiles with 128 processes on Hopper have been collected.
  
 
Those profiles are available here: [http://www.nic.uoregon.edu/~khuck/javaws/paraprof-super.jnlp ParaProf], [http://www.nic.uoregon.edu/~khuck/javaws/perfexplorer-super.jnlp PerfExplorer]. The client applications can only connect to the performance database from specific domains, and with authenticated access. Please contact [mailto:tau-bugs@cs.uoregon.edu the TAU team] to request access to the raw data.
 
Those profiles are available here: [http://www.nic.uoregon.edu/~khuck/javaws/paraprof-super.jnlp ParaProf], [http://www.nic.uoregon.edu/~khuck/javaws/perfexplorer-super.jnlp PerfExplorer]. The client applications can only connect to the performance database from specific domains, and with authenticated access. Please contact [mailto:tau-bugs@cs.uoregon.edu the TAU team] to request access to the raw data.

Revision as of 20:52, 2 November 2012


Overview

This is the TAU profiling MPAS-Ocean page.

MPAS-Ocean Sourceforge Page

The MPAS-Ocean code has been modified to use TAU as the timers, rather than the internal timers. This provides for both MPI performance measurement as well as PAPI counters.

The MPAS-Ocean developers have collected profiles on Hopper, with 192 to 16800 processes, using MPI only (no OpenMP yet). In addition, full callpath and communication matrix profiles with 128 processes on Hopper have been collected.

Those profiles are available here: ParaProf, PerfExplorer. The client applications can only connect to the performance database from specific domains, and with authenticated access. Please contact the TAU team to request access to the raw data.

Below is a brief analysis of the application performance.

Performance Analysis