In recent times, Apple had not submitted CPU performance results to the Standard Performance Evaluation Corporation, presumably because the G4 processors used in Macs lagged so far behind competing x86 processors. Apple pushed SPEC benchmarks when it started using the G5, but never formally submitted those benchmark results. Similarly, they have used SPEC to promote its new Intel Macs in marketing materials, but Apple has also recently started submitting SPEC benchmark results officially.
For marketing purposes, Apple uses the SPEC2000 benchmarks probably because that's what most of the world uses. However, Apple has only officially submitted results for the brand new SPEC CPU2006 series of benchmarks. Currently the only results available are for the previous iMac Core Duo 2.0 model, with SPECint_base200 = 10.1, and SPECfp_base2006 = 8.88. At least for integer performance, that puts the Intel Core Duo 2.0 in the ballpark of the 2.6 GHz AMD Athlon 64 FX-60, despite being a low power laptop oriented chip. (These Core Duo benchmarks are performed under Mac OS X, and use the latest Intel compilers. It should be noted however, that these Linux Athlon benchmarks are performed with gcc, which may be slower.)
It will be interesting to see how a 2.33 GHz Core 2 Duo Merom fares in these tests. My guess is that it should be close to 50% faster than Core Duo 2.0. That's what Apple claims for SPEC CPU2000 at least:
Subscribe to:
Post Comments (Atom)
1 comment:
Yes, Apple claims 40-50% added speed in that benchmark while increasing the clock rate of the processor by 16%. Not bad but not really spectacular either, especially as other benchmarks tend to give less of an increase and the processor generates as intense heat as yonah.
--J
Post a Comment