That's a nice OC cat. I paused for a second when I noticed the 2x4MB L2 cache. I'm just not used to shared L2 cache and was expecting 4xsomething. As I've said before shared cache currently boosts single threaded performance since that single core gets it all.
On the other hand the L3 cache of the phenom (which is going to grow to 6MB) can be modified by all 4 cores which is far superior but has yet to be taken advantage of. As more applications become multithreaded the value of shared L2 fades as cores have to fight for cache.
Anyone with a Q6600 @ 3.6GHz shouldn't really care. GPUs have a hell of a lot of catching up to do before that becomes a bottleneck.
