The BS cache hit efficiency on this thread is approaching 95%. :)
At this point, everything that has been said with respect to what AMD is going to do over the next few months is pure speculation. Since I also can't add anything meaningful there, I won't try. <g> I'll simply point out the following which I believe to be fairly reliable:
AMD will finish .18 micron with the fastest IPC processor (based on the leak that Palomino will top out at 16% faster than tbird - that would be around 1.7 or 1.8 GHz). Intel will finish .18 micron with the fastest MHz processor. It may just be me, but I'd be more proud of the former, than the latter. It is unknown if these cores will scale comparably at .13 micron, particularly with SOI in the mix.
AMD is going to try to overcome the perception that faster MHz equals a faster processor. We do not know for sure the particulars of this campaign. If done well, they'll have some degree of success, but if done poorly, I can't see it hurting them worse than the above perception itself (with respect to Joe Idiot consumer).
AMD has a major new processor core slated for .13 micron which could be big (but maybe not). Intel has no major new plans for an x86 core, so would have difficulty responding if Hammer does turn out to be big, SMT notwithstanding.
AMD should maintain IPC superiority (clock for clock) at .13 microns based on .18 micron performance, even if Hammer doesn't materialize. SOI will help keep them close to P4 scaling. Intel hopes their core scales well at .13, but it is an unknown. Hot spots have been a problem (albeit minor) for them at .18.
Predictions of wreck and ruin for AMD are a bit premature, IMHO. Actually, it's going to be a very interesting to see how this all unfolds.
Regards,
eCo |
