C Optimization, Pentium 4, 4 May 2004, tree-ssa/3.5/3.4/icc

[Scott Robert Ladd]

Diego Novillo wrote:
> Don't you mean, tree-ssa gets clobbered by ICC? 67 secs is a lot longer
> than 22 secs.

Excuse the loud profanity coming from the direction of Florida... :)

The headings are in the wrong order. It should be:

test    3.4.1   mainline  tree-ssa   ICC 8.0
----  --------  --------  --------  --------
alma    67.4      67.9      18.9      22.0
arco    25.4      26.1      26.1      20.8
  evo    52.8      53.4      53.2      37.9
  fft    32.0      32.0      32.1      34.5
huff    17.9      23.6      22.2      16.7
  lin    22.5      21.2      21.5      21.3
mat1     7.7       7.7       7.6       7.6
mole    75.2      73.9      75.6      12.1
tree    25.3      29.1      24.8      27.5
----  --------  --------  --------  --------
total  326.3     334.9     282.1     200.3

I reused a heading from a previous article; while I fixed the order on 
the Opteron article, I failed to do so when reporting Pentium 4 results.

Ugh. I need more coffee.


..Scott

-- 
Scott Robert Ladd
Coyote Gulch Productions (http://www.coyotegulch.com)
Software Invention for High-Performance Computing