Replies to post #67336 on Advanced Micro Devices Inc (AMD)

[chipdesigner]

Wildly overoptimistic, IMO. Then again, you did predict 4GHz+ DC P4 parts @ 65nm, so you're nothing if not consistent.

[HailMary]

Conroe is slightly constrained at 65W,

Why is Intel constraining their next gen desktop part to 65W? The competition is doing 90W in this arena. I understand the advantage of doing so in mobile and server, but desktop? It seems to get the highest performing gaming chip, they should open this up and see how high they can clock it. My thought on this is the new core is simply not power limited/constrained at all. At the highest frequency they can muster at the highest voltage their process can tolerate, it still uses less than 65W. That theory doesn't really explain 80W Woodcrest though.

I guess Intel is simply going for performance per watt, and not overall performance, or they simply hit one out of the park and achieved both all under 65W. If Intel didn't think they could have the highest performing desktop chip at 65W TDP, they would have specified it at a higher TDP, since nobody in the industry is really pushing for a 65W TDP for desktop. 90W with a lower power idle is sufficient. That is what makes me wonder about my AMD investment. The specs seem to lead to Intel confidence in their next gen cores.