Replies to post #40553 on Advanced Micro Devices Inc (AMD)

[KeithDust2000]

BUGGI, it sure is, many people have waited for those new Pathscale results, and for good reason.
Absolutely killer scores.
INTEL has had the advantage of their special compilers for years, AMD gets them for the first time now. And I´m sure this is only the beginning...

[KeithDust2000]

BUGGI, don´t worry, AMD tanks regardless of news ;-)

[wbmw]

Re: SPECfp_rate2000 = 70.0

That's a pretty substantial improvement over the previous Opteron 850 score.

http://www.spec.org/cpu2000/results/res2004q2/cpu2000-20040503-03006.html

(Differences in bold)

Operating System: SuSE Linux 8.0 SLES 64 bit (SP3)
SuSE NUMA kernel 2.4.21-143
Compiler: PGI Fortran 5.1-3
SuSE gcc33 optional compiler (from SLES8 SP3)
AMD Core Math Library (Version 1.5) for AMD64
Base: 48.5
Peak: 49.8

Operating System: SuSE Linux 8.0 SLES 64 bit (SP3)
Compiler: PathScale EKO Compiler Suite, Release 1.1
SuSE optional gcc 3.3 (from SLES8 SP3)
PGI Fortran 5.2 (build 5.2-0E)
AMD Core Math Library (Version 2.0) for AMD64
Base: 61.5 (+27%)
Peak: 70.0 (+41%)