Profits. Overclocking allows people to buy cheaper processors and modify them to perform like their expensive counterparts. When I bought my Q6600 there were two Core2 Q series processors out and the only difference between the $280 and $1100 models was .266GHz. Behind the scenes the low-end had a multiplier of 18 and the high-end had 20 but in essenence all you had to do was overclock the FSB from 133MHz to 145MHz and you just saved yourself $800. In the past, the money you just saved was money Intel wasn’t getting. Now it is. Intel’s standard line of Sandy Bridge processors switch from overclocking to “turbo boost”. It overclocks the CPU but also disables cores so you’re essentially not getting any extra power at all. They have an overclockable series that costs more and gives a little more flexibility with the “turbo boost” settings but overclocking as we know it is no longer an option on Intel processors and likely never will be again.
I called this, btw, as soon as I heard that Intel was locking their core i7 processors to intel-only chipsets. They called it a design imperative, I called it profit-driven. AMD managed to put the memory controller on the CPU without locking out 3rd party chipset manufacturers, there’s no reason Intel couldn’t have done the same.
Were I to build a new computer today it would be a last-gen Bloomfield Core i7 but I’m happy with my 4 year old Q6600 and will be for the foreseeable future. I’m also extremely fond of my oldschool nVidia nforce chipset motherboard, you just can’t get those like you used to. You know I can actually rebuild RAID arrays while windows is running? When my RAID5 dropped a drive I was able to rebuild the entire array while playing L4D2 from the array that was being rebuilt. It was awesome. Intel’s RAID controllers are, well, horrible on the best of days. You can’t even build a mirror without wiping both drives. Terrible.