It’s been almost ten years since Intel put a “multiplier lock” on its Pentium II models.
About a year from now, the other shoe will drop.
When the cheaper Nehalems come out, the socket 1160 kind, they’ll contain additional circuitry to prevent FSB overclocking.
To overclock a Nehalem, you’ll need to buy the upper-end, socket 1366 Bloomfield models, which are shaping up to be an expensive proposition. The lowest priced quad Bloomfield is supposed to be priced around $316; at least the initial motherboards are also likely to cost a ton, and three-channel memory isn’t exactly a cost-cutter.
This is what people do when they don’t have real competition.
It is also foolish. Of course, you’d expect me to say that, but hear me out.
Right now, Intel dominates the O/C world because its processors when overclocked provide a lot more performance than AMD for not too much more money.
The key phrase is “not too much more money.” Add another one or two hundred dollars to the price tag and “not too much more money,” turns into “too much money” for too many people.
The people who are willing to spend the money are going to buy Bloomfields, which are going to be out long before the cut-rate stuff. The potential O/C market for the lower-end Nehalems are pretty much those who won’t spend the big bucks. Some will bite the bullet, but most won’t.
No matter how good Nehalem is, a sizable chunk of overclockers just won’t pay a lot for a system, and a lot more of them will put price over performance and go with AMD, no matter how relatively bad it is.
This Intel move, if true, would single-handedly revive AMD’s share of the O/C market. Granted, this won’t exactly make AMD profitable, but if I’m Intel, I wouldn’t give Green squat.
AMD could greatly encourage this trend by saying something about the compatibility of its high-K 45nm and/or Bulldozer chips with AMD motherboards. People would be more likely to settle for a bit less initially if they know they can upgrade later.