How well are you expecting the early Prescotts to OC?

[AgaBoogaBoo]

I think they'll OC somewhat better because of the smaller process they're made on, but the amount of power they need is also higher and thats why some motherboard companies are wondering if it will work on all current 865 and 875 boards. I get the feeling the early ones will OC really high past 4Ghz without much trouble unless heat is also very high.

On the side of motherboards, do you guys think that all of the current 865 and 875 motherboards will be affected or just certain ones that can't provide enough power? I'm guessing some of the better overclocking motherboards will be able to provide the power it needs, but I'm not sure. Do you have any idea?

 

[crims0n]

What are you talking about? The micron process has nothing to do with how well you can overclock, its to do with the quality of the transistors in the CPU core and how much heat is generated during their changing of state, the lower quality the more heat generated and less stable overclock. The higher quality the better clock speeds obtainable, this is also tied in with CPU pipelines, cache and instructions per clock.

Of coruse the lower end early stage chips are going to overclock more - because they will have more headroom. As far as I was aware Socket 478 is gonig to be discontinued and replaced with 775, though I am unsure if an earlier stage of Prescotts will work in our current 875/865? motherboards.

http://www.tomshardware.com/cpu/20030623/p4_3200-03.html

 

[dexvx]

Early Prescotts will work with good quality 875/865 boards that are officially sanctioned to run Prescotts. They will not work with low end 875/865 boards.

 

[Serp86]

the new abit ic7 max 3 says prescotts are guaranteed to run on it.

 

[ViRGE]

Honestly, I don't expect the Prescott to overclock well. Judging by the numbers we're seeing for heat, coupled with the smaller die size, I think cooling is going to be too large of an issue to get a great overclock out of. I'm expecting something similar to the original P4 introduction.

Edit: As Wingnutz pointed out to me Prescott will have 1MB L2, so the die size won't shrink like what we saw with the Palamino-TBred. Heat could still be an issue 'tho

 

[zephyrprime]

I don't expect it to overclock well. Early processes never do. Also, we can expect there to be nothing but high end prescotts when they first come out so each will actually be running at the max of their abilities. (unlike the athlon 1700+ which only runs as slow as it does because AMD is filling a market niche with it.)

What creates good overclockers is mature processes combined with a need for low end products.

 

[dexvx]

Originally posted by: zephyrprime
I don't expect it to overclock well. Early processes never do. Also, we can expect there to be nothing but high end prescotts when they first come out so each will actually be running at the max of their abilities.

Let me think back. Northwood, 1.6a launch which came shortly after the 2.0 and 2.2. The initial batches ran 2133 no problem, and many ran as high as 2.4.

 

[Jeff7181]

I think it's rare that the first batch will overclock well. It all depends on how far Intel has been able to get with their new manufacuring process... if they're only able to make 3.4 and 3.6 Ghz processors right now, don't expect to be able to overclock beyond that. However, if they have some running at 4 Ghz fairly easily already, expect good overclocking results.

I guess the shortened version of that is, "wait and see"

 

[Afro000Dude]

I think i read somewhere that the strained silicon that will be used in the core does not react well to over-volting, which would further limit overclockability. I think it was an Anandtech article...