...
To understand the impact of I/O on the model performance start by taking a look at the time taken by the PIO functions (grep for pio*) in the timing subdirectory of the job run directory. (At NERSC, ALCF, and OLCF, these data are also archived in the performance_archive . Bad for each system. Site is system specific.) Bad PIO settings can cause the model to crash, or have horribly slow output performance. But you want to make sure you have a reasonable processor layout before calibrating the PIO settings.
...