3.7 Parallel Programming Tips

• Performance on an SMP system may be limited by memory bandwidth. To test for a memory bandwidth limitation, run multiple copies of a program as a single processor application on an SMP system. It should take the same time for 2 independent copies to run on 2 processors as it takes for 1 copy to run on 1 processor. If 2 copies take longer to run, your system probably has a memory bandwidth bottleneck.

• Use the OpenVMS Alpha routines lib$init_timer and lib$show_timer to view the real elapsed time of code sections that run in less than 1 second.

• When you use PCF directives and you do not specify the KAP /concurrentize qualifier, you must add the OpenVMS Alpha version-specific linker qualifiers to the kfort command line.

• Experiment with each of the following qualifiers to determine if they improve the run-time of your program.
  /fuse
  /fuselevel=1
  /ipa
  /ipa_from_files=<file>,<file>
  /ipa_optimize=2

  However, the above qualifiers may increase the amount of time and memory KAP needs to process your source files.