You just proved my point. Thank you.
The idea with the 45nm shrink is the same as the 65nm compared to the old 90nm. You shrink the size of the transistors, this allows you to pack more and increase the number of overall transistors. This then increases the speed at which data is processed, even when it is running at the same base speed (3.2 GHz Pentium 4 Prescott vs 3.0 GHz Core 2 Duo E6850?) The idea here is there is more data processed at once.
God how do I explain this. OK, you have a highway. It is a certain width in feet, let's say 40 feet wide. The 65nm core on this highway is using 4 lanes, 10 feet wide each lane. Traffic is flowing at 60mph. Now, if you shrink the size of the lanes from 10 feet each down to 6.6 feet wide each, you can now fit 6 lanes onto this highway. Flow of traffic has now increased overall. the speed is the same, but more cars are traveling on the highway.
Same thing here. Even at the same speed, more data is being processed per second thanks to the smaller lanes.
Comparing a Q6600 to an E8400 for gaming is just silly. Until the majority of games find some benefit with quads, the 45nm duals will be the winners in this particular match.
Did you just make this up?!? that is completely wrong... first if you are running the same speed and same multi the new 45nm is only 2-5% faster PERIOD. The only advantage 45nm has is it can run on less voltage which in turn emits less heat then you can run at higher speeds but as for this lanes switch...highway.. traffic analogy is very inaccurate.
EDIT: also until they bring out a revision to any of the 45nm chips it is like playing with glass... so watch out