When it’s clock-throttled.
See Mike Magee of the Register (hope you’re feeling better) explain this.
See Intel scream.
See AMD stay quiet, because they’ll have much the same thing shortly.
The main issue is how often clock-throttling happens. If this just occurs under battery power, or under extreme heat conditions (as Intel claims), that’s one thing. If it effectively happens all the time (which is the contention), then why are you paying big bucks for a 1Ghz CPU when it never does any serious work at 1Ghz?
Nor is this necessarily for notebooks only.
It wouldn’t surprise me if this concept gets moved to the desktop someday. With clock-throttling technology, CPU manufacturers could give their chips higher desktop Mhz ratings than they can now, and just throttle back when the CPU was in danger of doing any real work at that speed.
Don’t laugh, that’s just what AMD could do with Palominos equipped with PowerOn. Not saying at all that they will or might on the desktop. I just hope I didn’t give marketing people any ideas. 🙁
What does this have to do with overclocking?
There’s some hints that possibly there is a way to override the Intel multiplier lock. There’s a number of reasons why this wouldn’t work, and at the very least, we’re talking mega BIOS hacking, but finding out just how Intel does this would be an interesting first step.
Verifying whether or not this happens seems simple enough; just run a CPU benchmark test like Sandra or CPUMark along after running something like Prime95 or Content Creation for a little while on one of these 1Ghz notebooks, and see what you get. You’d also want to run Intel’s own handy Processor Frequency ID Utility to verify the speed.
To do this, of course, you do need to be able to test a $3,000 notebook for a couple hours.
Is there anybody who works in a computer store who could run a couple tests on one of these 1Ghz notebooks and get some screen shots? If you can, and you’re game, please write me.