Going from 65 to 45nm may mean the main difference in top oc clocks, but even at both at 45nm, pens slightly outclock the i7s. Plus, they dont heat up, and thats just part of the native quads makeup, with everything on board 1 chip, like AMD has been doing for awhile. Having SMT effects all this, as well as a IMC structure. Really, whats amazing here is P2 doesnt get near as hot as i7, being so similar, tho, at same loads, we will have to see, once Denebs released, if at the same load, i7 gets hot or not
I think you don't realize how much power dissipation is due to forcing work through your cores at twice the normal rate.
Core Temp readings for my i7 (3.34 GHz core, 3.6 GHz uncore, stock HSF, PWM auto-regulated)
68-70 C, full load, HT on
62-64 C, full load, HT off
44-46 C, idle
The differences would be more dramatic with a constant fan speed.
Realize that a Q6600 @ 3.34 GHz, using a very similar HSF, with no option of HT, loads up at around 70C.
While I won't say that a PII can't beat my power consumption at 4 threads, I'm almost certain that if everyone is allowed to fine tune voltages (thereby turning off CnQ/Speedstep) and adjust clocks to a minimum performance requirement, the i7 would be much, much harder to beat, as even the Yorkfields lose to this "130w TDP" CPU.