# $Id$

The programs in this directory are designed to measure the I/O performance for
various of APIs as well as access patterns.

C/aggregation.c
   o This program writes a series of 2D variables with data partitioning
     patterns of block-block, *-cyclic, block-*, and *-block, round-robinly.
     The block-* partitioning case writes 1st half followed by 2nd half. The
     same partitioning patterns are used for read. In both cases, nonblocking
     APIs are used to evaluate the performance.
   o Parameters:
         * NVARS: a defined C macro, the number of variables
         * len:   dimension size of local variables, len x len
   o Write and read performance are measured and reported separately.


C/write_block_read_column.c
   o This program writes a series of 2D variables partitioned in a block-block
     pattern into a new file. The file is re-opened to read all the 2D variables
     but in a 2D *-block pattern.
   o Write and read performance are measured and reported separately.


FLASH
   o This benchmark is algorithmically identical to the FLASH-IO kernel.
     FLASH is a reacting hydrodynamics code developed at University of Chicago.
     http://flash.uchicago.edu
   o This distribution contains only PnetCDF I/O method.
