CPU Life Expectancy

[clarkey01]

See a lot of post about overcloking has really got me in the mood, seeing Duvie and Zebos efforts has persuaded me to have a go, but Id like to know the following :

When you overclock, how does this shorten the life of the CPU ? whats different (besides from the multiplier) from the CPU being set at that speed in the fabs >?

I am going to wait till march-ish, to see the second wave of 90 nm of athlon 64s, and with better yields 3 Ghz maybe attainable (wink wink wingznut).

Cheers

 

[Zebo]

negligible. CPU's made to last 100,000 hours. that's like 15 years on all-the-time, your CPU will be history by time long before that. little more vcore and high current ain't going to shorten that much. Go safe and slow the first time like all vigins should and you'll have zero problems.

stay within the realm of highest rated chip in iteration.
In the past most people took cheap $80 2500 bartons left stock vcore and simply bumped them up to $200 3200 levels. It's the same chip after all.

Now a good starting point may be a 2800 to 3400... a presscot 3.0 to 3.8...alternativly a 939 3000 to 3500. Then, I think, you're not even exceeding maufactures MTBF.

 

[superkdogg]

Good advice, Z. You could even go a little further, but if you're worried about safety-Zebo's got it. On the other hand, when you feed more vCore you can often get more than the highest rated chip, but then you shorten life expectancy somewhat. As previously noted, the chip would still almost certainly last until you wanted to upgrade anyway. The major exception to this (other than bad luck) would be if you have poor cooling. High temps over a long time would do more damage than higher vCore with good cooling IMHO.