Overclocking LGA 775 3.0

[nissmo300]

Since newegg now carries the "Intel LGA775 530 3.0" what do you think it can OC to?

Do you think it'll be better the a 3.0C? 3.0E? Both?

[Zebbo]

Well, don't know about 530 LGA P4 3.0 though, but those 550 CPU's has gone pretty well. With default heatsink + Delta 80x80 modded to it straight to +4GHz / 1.4V. 540 CPU's took like ~3.95 / 1.55V.

[Albuquerque]

1.55v?!? Isn't the stock voltage on these like 1.3v? I would assume that's only for extreme benchmark runs and not any sort of normal day-to-day activity...

[batboy]

Default Prescott vcore is usually 1.3875v, but I've seen a few at one notch lower and a few at 1.4v. Yeah, 1.55v is getting rather high. But, I think the CPU can handle it as look as it's not overheating. With socket 478 mobos, high voltage can kill the mosfets. But, with the i915/i925 chipset mobos, the power circuitry is far more robust, so in theory anyway, they should handle higher vcores much better.

[Albuquerque]

Even at 1.3875vcore stock, you're talking about an almost 15% overvolt. When was the last time you suggested someone run a P4C at 1.75v for extended periods of time? Maybe the better motherboards can handle the overvoltage, but I'm not convinced the processors can.

A 130nm-process Pentium 4 is half-expected to die of SNDS at a 15% overvoltage, why would we be suggesting this kind of percentage on an even smaller 90nm process?

[batboy]

When I fried my IC7, I was running a Prescott 2.8E at 1.6v on a suicide run for a max O/C (got a screen shot at 3.95 gig just before I toasted the mobo). The CPU was fine and later went on to do 4.1 gig in Deathstar's Prommy. But, you're right about that vcore being a little high. I've been recommending people to limit Prescott vcore to 1.5v for long term usage.