Table of Contents
This appendix presents the usage display from each
benchmark program. A benchmark prints the display when invoked with no
parameters or with the single parameter -0.
This appendix presents the displays from the CELL benchmarks.
fastconv -- fast convolution benchmark for Cell BE
Sweeping pulse size:
-1 -- IP, native complex, distributed, single-expr
-2 -- IP, native complex, distributed, Fastconv object
-11 -- OP, inter complex, non-dist
-12 -- IP, inter complex, non-dist, single FC
-13 -- IP, inter complex, non-dist, multi FC
-21 -- OP, split complex, non-dist
-22 -- IP, split complex, non-dist
-23 -- IP, split complex, non-dist, multi FC
-32 -- Multiple coeff vectors in time domain, IP, native complex, non-dist, single FC
-42 -- Multiple coeff vectors in freq domain, IP, native complex, non-dist, single FC