Re: New record for spec fp peak on x86

Greg Lindahl wrote:
http://www.spec.org/cpu2000/results/res2006q2/cpu2000-20060612-06166.asc

is an HP submission on a Woodcrest system scoring 3049 on SPEC fp peak
-- first x86 score above 3000. It uses the PathScale compiler for 6
of 14 codes and those Intel compilers for the rest.

Interestingly, this 3049 score is higher than the best Itanium2 score
I could find (3017, http://www.spec.org/cpu2000/results/res2006q2/cpu2000-20060613-06201.asc.)

My interest has been piqued by the alleged difficulty that Woodcrest
and
Conroe systems have doing RAID 5 (see the Inquirer). Does anyone have
any further insights into that story ?

Cheers,
Rupert

.