Tony,
Some are being objective (not to categorize this poster as an AMDroid, but most of his posts are on AMD threads).
I have not seen a single person on Intel thread express an objective assessment of P4. Anyway you look at what the processor delivers today, you must conclude that it is a disappointment.
I personally expected the outline of the story to be something like Pentium II, with weak initial benchmarks (the one on Tom's site was the best example), but a substantial jump in performance of the production systems.
It didn't happen. The weak benchmarks that have been leaked over last few months were confirmed by the production systems.
While the processor may amount to more in the future if Intel guessed correctly the direction where the software will move, I find this bet to be a long shot. How much did things change in say last 4 years, which is a typical life of 1 generation of a microprocessor? Not much. I am running pretty much the same software, maybe version 8 vs. version 3 that I did 4 years ago, but the mix didn't change much.
One thing that is emerging is that Intel will have to maintain clock speed lead of 30% or more just to keep up with performance of current AMD processor, which means Intel will have to have a 2 GHz processor to beat AMD 1.5 GHz processor, a 2.6 GHz P4 to beat 2 GHz Athlon.
Going forward, the roadmaps of the core changes of Intel and AMD seem to be:
Intel AMD
Q1 01 Mustang core tweak
Q2 01
Q3 01
Q4 01 Northwood shrink Thoroughbred shrink
Q1 02 New Hammer core
It seems obvious that Intel again failed to "slam the door" and the race will at best be too close to call for the next year.
Not much is known about AMD Hammer processor family, but there is a risk (for Intel) that AMD will leap forward, and Intel will be stuck in the same situation as we have seen throughout the last 12 months, as Intel was just barely holding on with the Coppermine vs. Athlon, facing unfavorable odds.
Joe |
