Sandy Bridge Processor Degradation

[Grooveriding]

Help this layman out with some of this discussion.

I'm going to put my 3930k under water and crank the voltage up to 1.45-1.5V and run it at 5ghz. I've confirmed I can do 5, but temps are astronomically high with my NH-D14.

So, is it the voltage alone that causes degradation, or does it happen in tandem with heat.

All I care about is getting about 18 more months out of this CPU if that matters. I'm assuming I should have no worries

 

[Smartazz]

Help this layman out with some of this discussion.

I'm going to put my 3930k under water and crank the voltage up to 1.45-1.5V and run it at 5ghz. I've confirmed I can do 5, but temps are astronomically high with my NH-D14.

So, is it the voltage alone that causes degradation, or does it happen in tandem with heat.

All I care about is getting about 18 more months out of this CPU if that matters. I'm assuming I should have no worries

I wouldn't go over 1.4v personally as that would probably offer good OC results. I remember hearing Intel said the recommended max voltage for SB-E is 1.4v.

 

[Hugo Drax]

I don't remember where I saw it but I think that the MTBF (Mean Time Between Failures) on our processors is something like 1,250,000 hours or somewhere over 146 years. So while running extreme in tempuratures or high voltages may shorten the life of your processor I would guess that the temps and voltages you posted wouldnt shorten it to the level that you see the difference.

I wonder what kind of CPU intel will offer in a 146 years from now.

 

[Smartazz]

I wonder what kind of CPU intel will offer in a 146 years from now.

I wonder if CPUs will still be in use on a large scale in 146 years. In the past 146 years we've gone through punchcards, relay devices, vacuum tubes, transistors and we're now on integrated circuits. Imagine what we might have if we have another four paradigms.

 

[NumericalMethods]

Help this layman out with some of this discussion.

snip

Groove, this is the link you need http://www.anandtech.com/show/2468/6

It is the picture I hold in my mind when talking about overclocking all the time. The reason waterers go higher with their voltage is simply due to the fact that by maintaining lower operating temperatures they can reduce the processor degradation rate.

Also bear in mind, in general, you need less voltage for stable operation at lower temperatures. With a water cooled setup you will likely need less voltage to hit your 5 GHz target, providing the loop does reduce temperatures over the Noctua.

 

[BrightCandle]

Intel's max voltage for SB is 1.35V, it was 1.4V for 45nm chips. Both responses (temperature and voltage) are not linear so the extra 10% voltage from stock design is more significant than the 10C you'll drop via watercooling.

If you are really talking about 1.5V for 18 months you better be talking about watercooling with the ability to maintain <5C delta at around 300W and ideally at below ambient (chilled). That is one impressive cooling system btw, but you are going to need every degree of drop you can get because SB's have died in overclocking at just 1.45V. Many 1.5V runs are either suicide benchmarks or under phase cooling.

 

[Quantos]

[...] "there isn't enough data out there to answer the question properly" [...]

Surely this has to be more simply than maximum entropy?

:biggrin:

 

[Smartazz]

Voltage increases power consumption in an exponential fashion, so I'd imagine that 1.35v-1.4v is a much bigger jump than 1.3v-1.35v. I'd be happy with a solid two years of lifetime out of my 2500K, but I don't want to sell a degraded processor to somebody. My overvolted Q6600 and Athlon X2 barely degraded, even to this day, but I was relatively tame with their voltages.

 

[ed29a]

No need to worry about processor degradation, pay 20$ extra to Intel for insurance and you are good to go.

 

[PlasmaBomb]

I don't remember where I saw it but I think that the MTBF (Mean Time Between Failures) on our processors is something like 1,250,000 hours or somewhere over 146 years. So while running extreme in tempuratures or high voltages may shorten the life of your processor I would guess that the temps and voltages you posted wouldnt shorten it to the level that you see the difference.

MTBF, you are doing it wrong.

By the way 1,250,000 hours is 142.6 years (1 dp), so Math fail as well.

 

[Grooveriding]

No need to worry about processor degradation, pay 20$ extra to Intel for insurance and you are good to go.

This is unbelievably amazing. Thank you, sir!

 

[Kenmitch]

No need to worry about processor degradation, pay 20$ extra to Intel for insurance and you are good to go.

Reasonable pricing and it states that you can transfer the warranty to another person also

Hmm....My 2700k will do 5ghz around 1.45v's or so. Keeping it cool enough may be an issue under heavy loads.

 

[Puppies04]

No need to worry about processor degradation, pay 20$ extra to Intel for insurance and you are good to go.

I didn't even know this existed, seems comically underpriced... when did they start doing this?

 

[Smartazz]

Isn't the likelihood of a total failure very low in comparison to the chips that simple require more voltage to maintain their overclocks?