Ever since Intel’s Prescott was introduced to the world, our opinion on buying Intel processors has been quite simple.
No. Just no. They were literally too hot for the average overclocker to handle, and the results they produced for the average overclocker weren’t worth the bother.
There are now some early overclocking results on the first generation of 65nm processors, and they aren’t bad.
Even assuming the architecture of these chips is no better than Prescott, 4.5GHz for the single Cedar Mill should be pretty competitive against current Hammers overclocked the same way.
4.25 for the dually is a little more questionable, but Presler ought to be a bit better than the slapped-together Smithfield, so it will probably be competitive, too.
While these look to be hotter chips than current Hammers, heat levels are rather better than Prescott and ought to be tolerable.
I’m not trying to say these are great chips, or even “normal” chips by historical standards, but historical standards are just that in this era, historical. We can’t expect huge speed increases from a process shrink any longer.
Nor am I saying that these chips will become the overclocking chips of choice early next year. AMD may well be able to tweak 90nm a bit more when the socket M2 incarnation comes out.
Perhaps the example of pro wrestling can help us here. During the Prescott era, Intel was essentially tossed outside the ring for all but the hellbent (figuratively and literally).
Cedar Mill and Presler basically represent Intel crawling back into the overclocking ring. This doesn’t mean they’re going to win the match, that will be up to its tag team descendant-of-Pentium-M partner later in the year
But at least they’ll be back in the ring.