some benchmarks

[Steve Kargl]

On Sun, Nov 27, 2005 at 01:30:55PM -0500, Jack Howarth wrote:
>     Since gcc 4.1 is building on Darwin again, I decided to benchmark
> it performance at -O3 in the xplor-nih testsuite versus the current
> Fortran XL compiler with full -O5 optimizations. Here are the results
> which are a tad discouraging.
> 

I've observed that -O3 is sometimes (often?) not a good
choice.  I'd suggest -O3 -funroll-loops -march=whatever_cpu.

A more important measure is whether gfortran is producing
the right answer. ;-)

> dsearch.........ok   4.7100 	4.2939

We did better on 1 test.  Do you know if xplor-nih
use dot_product and matmul?  Recent discussions
suggest that these are suboptimal.

-- 
Steve