Profiling High Performance Fortran (HPF) programs is accomplished through parallel runtime data recording and post-execution analysis with the pprof utility.
Parallel profiling provides analysis of HPF programs executing in the Digital Parallel Software Environment (PSE). The information in pprof reports is used to locate hotspots and to identify aspects of program execution that might benefit from optimization. The profiler also provides system run-time performance statistics for processes running under PSE.
This chapter describes how to use parallel profiling. It introduces performance issues that are considered when designing a parallel profiling strategy.
The chapter is divided into two sections:
