Originally posted by: CrawlingEye
Originally posted by: Rainsford
Um, and about specs. Look at the specs for the P4. Looks good, right? Then why is it only saved by high clockspeeds? According to those specs you posted, the P4 should whoop the Athlon XP even at the same clockspeed. They sound great, don't they? But the P4 looses very badly at the same clockspeed as the Athlon XP (none of this SSE2 stuff since I don't see programs supporting it popping out of the walls). Now how would anyone know that from those great sounding features you posted if neither of these products were released? Specs can be made to "prove" anything.
For someone claiming to know so much about architecture, you've yet to hear of IPC yet, haven't you?
The AXP's have faster ALU's and a shorter pipeline, which limits their scalability (as well as how well they OC).
The smartest thing AMD's doing with the Hammer is putting the heatspreader on it.
I do not claim to know so much about architecture, stop putting words in my mouth. But I know exactly what IPC is, and I know that the AXP's is a lot higher than the P4's. That is why I said the P4 is only saved by high clockspeeds. My point is that without posting the IPC of the P4, it would be very easy to use some of the P4 specs to make it look super fantastic when compared clock for clock to an AXP. That is why you cannot look at specs of yet to be released processors as if they actually mean anything about real performance. Specs can be twisted, good benchmarks (and real use by real people) can't.
Heatspreader on the Hammer is a good idea to prevent some people from chipping the cores, and that's it. The heatspreader does not help cooling (at least I don't see how it possibly could). It is only protection against chipping the cores, something you should never do if you take your time putting on a heatsink.
Edit 2: Eh, nevermind