Performance Ratings
All this [*yawn*] arguing about the performance ratings has reminded me of something I was thinking about AMD marketing. Basically, the switch to performance-based model numbers killed the Duron and are complicating the introduction of the Hammer. To give Intel some credit, no reasonably knowledgeable consumer will expect a 2g Celeron to perform as well as a 2g P4 (now, whether they expect it to perform as badly as it does is another story). The two product lines maintain some independence from each other. Now that AMD has the PRs, they can no longer do this. An hypothetical 2000+ Duron should not underperform a 2000+ AthlonXP, otherwise it would not deserve its rating.
Ok, so far old news, I know. But, if September's A64s have the same or similar PR scheme as the AXPs, then AMD will need them to rank credibly higher than the XPs. I think this is a big part of the reason why current Athlons seem to be drastically underclocked acc'g to current reports, and it's also probably why they seem to be vacillating about releasing 400m fsb, barton 3200+, etc.: they need the A64 PR to exceed AXP PR, and they are not yet sure exactly how high they can get clock the launch-A64s. They need to keep a safe margin to insure that the A64 takes the unambiguous prestige spot in their product mix.
If this theory has anything to it, then it will be a very good sign if higher PR Bartons get released (for once, it doesn't seem to be a technical limitation!) because it would indicate that A64s are performing a good cut above that level.
On the flip side, a red flag would be an announcement or some credible rumors that AMD is switching to an Opteron-like system for naming the A64s, since it could be an admission that they could not compete performance-wise.
It suppose it all boils down to how well SOI is going. I'll be watching the Opteron release to find out.
