osbert :
sorry to hijack this thread but i have a similar question...
I've clocked my q6600 G0 to 3.2 on air easily... apart from the 12mb vs 8mb L2 cache on the QX9770 would there be any other performance differences? Would the difference from a 45nm process actually mean anything in real world performance?
I'm running at 400Mhz FSB which matches the QX...
cheers
os
Penryn has a few extra tricks up its sleeve:
Lower power consumption (both as a result of the shrink from 65nm to 45nmm and some other changes) mark it as a slightly better cpu ... but the real world difference is scarcely better than 3 - 5 % for a gamer.
Some sites show up to 10% performance increase (at the same clock speed).
Getting the Q6600 to 3.2 on air simply by selecting the 1066 FSB and putting on a decent cooler is a good result ... if your running 1600 then I am just scratching my head over that one ?? I assumed you dropped the multiplier ... or something aint right ...
Instead of running 9 X 266 (1066) you simply need to run 9 X 333 (1333) = 3.06Ghz.
If your running 400 (1600FSB) and the multi is on 9 then you have 3.6Ghz ... doable but she will get damned hot without watercooling I imagine ... or have you got it on 400 X 8 ?? ... much easier and the RAM gets a healthy bandwidth boost ... not that the core2 CPU's benefit much from that as the prefetch / cache is very good to start with.
I'd be more interested in stuffing the beast full of RAM, sticking 2 cheap large HDD's in RAID0m and putting the best graphics card in it you can afford.
A quad at just over 3Ghz is plenty ... it won't be cpu limiting any games anytime soon ... not with a single GPU and 22" monitor.
Anyway ... hope this helps ... sorry for the late reply ... Christmas etc etc.
Good luck.