> Re: "Intel's normalized scores are much more optimistic, claiming that the 1.40GHz P4 slightly outperforms the
> 1.00GHz PIII[in SYSmark2000]"
> I don't believe Intel has ever had this claim publicly, at least I haven't seen it. Can you point us to where you saw it?
You are absolutely correct when you say that Intel has not publicly made this claim. I'm mildly ashamed that I'm getting ahead of myself here. I completely confess that I cannot substantiate this, though in my mind it is a near absolute certainty (about as absolute a certainty, for example, as my suspicion that AMD will announce a chipset on Monday).
I have been, in this case, contacted by two people with scores that they claim are benchmarks from Intel Confidential documents. These two people are completely independent of each other and neither of them has ever rubbed me wrong in the past. One of them has benchmarks normalized to PIII 800EB, and I believe the other has scores normalized to a PIII 1000B (not entirely sure, he may have 800EB-normalized benchmarks which also include PIII-1000B scores normalized to the 800EB, and may have sent me renormalized scores, that sort of thing). The information from the two people is otherwise pretty much identical, and I trust them.
I know that this isn't nearly enough for you to have anything but skepticism for these numbers. Nonetheless, I am confident enough to use them in mixed company.
-JC |
