Long term negative effects of undervolting

[RejZoR]

I'm running my Core i7 920 at 3,4GHz using 1,15V (QPI at 1,25V). And i think it could go even lower. But i'm wondering if this could have a negative effect on my CPU. I know that too high voltages can damage CPU, but what about lower ones? I'm not geting any BSOD's or crashes. Anyone with electro tehnic knowledge who could 100% confirm or deny this? I mean i heard that in some cases when you underclock and/or undervolt, the chip starts to pump abnormal amps which then kill it later. Could this be the case?
I'm undervolting as much as possible because heat is a problem and this chip doesn't seem to require much voltage to work properly.

[Snow Crash]

subscribed ... same heat frustration running 3.2 GHz @ bios 1.125 / under load 1.1 w/1.2125vTT

[nicspits]

it's true a chip will draw more current when it's being used at a lower voltage than a higher one, but not by much. It's kind of like an engine taking more fuel when you step on the gas, but if you have bigger injectors, well, depends on how much you step on the gas right. so assuming you're gaming and using 90 percent or so, just don't try and force it so low that it's unstable and you should be fine. The heat is much more likely to degrade the life of your cpu but either way, we're talking small percentages of lifespan. It's only when you get 30-40% out of spec that it really starts to cut the life quick, and that's more with overvolting than under.

See if 1.1V runs it stable with decent temps under orthos for a day, if so, leave it there.

[sunny7day]

I think undervolting won't damage a cpu. But overclocking will, even without overvolting.

[Peen]

I killed an Athlon 64 3700 once from undervolting. Never thought I would see that happen

[unclewebb]

I nuked Windows once by under volting. It did a BSOD about half way through booting up and the easiest fix was to do a complete install. I tend not to forget when I do something really dumb like that.

Go down in small steps and make sure it's stable at each step before going lower.

[tiro_uspsss]

I aint knowledgeable, but I cant see why under-V would hurt CPU

[dengyong]

I've been running an e8400 under 1v for about two years with no ill effect.

[ThugsRook]

I've been running an e8400 under 1v for about two years with no ill effect.

e8200 @ 0.82v here for a year already, not a single problem.

[3NZ0]

Interesting thread, wouldn't mind seeing some good input on this matter.

I run my bedroom machine on a 920 c0 @ 3.5ghz w/ ht & turbo @ 1.17v / 1.25v vtt + all c states to keep temps and noise to a minimum (machine is left on 24/7 unlike my work machine)

I've been wondering if it could be a problem and I should just leave the volts at stock?

[donmarkoni]

I was researching on this and talked with my uncle who is a physics professor and came to a conclusion:
ACCORDING TO OHM'S LAW ( I[amps] = V[volts] / R[ohms] ) UNDERVOLTING PRODUCES LESS CURRENT IF RESISTANCE IS CONSTANT. So... I answered you nothing.
Problem is that the semiconductors does not have constant resistance (they have dynamically changing resistance) and there is also temperature implicated change in resistance.
I haven't draw any definite conclusion yet, but IMO under "normal" operating conditions you should be fine. If you stress your undervolted CPU 24/7, maybe you could kill it with high current due to high temperatures. But what is undervolting by Intel's specs, below 0.85, 0.80V for i7 for example?
I'm very curious about that dead Athlon64 3700. What were the operating conditions?
If anyone cares to read, here are the links to some material used for brain-gymnastics.
http://en.wikipedia.org/wiki/Ohm's_law
http://www.allaboutcircuits.com/vol_1/chpt_12/6.html
http://en.wikipedia.org/wiki/Temperature_coefficient
http://en.wikipedia.org/wiki/Semiconductor
http://en.wikipedia.org/wiki/Electrical_resistance
http://en.wikipedia.org/wiki/MOSFET
About PSUs:
http://www.marcspages.co.uk/pq/2410.htm
Should be nice to find something about PWMIC. Maybe similar could be applied to see if it could kill the CPU.
Last, but not the least, discussion:
http://www.electro-tech-online.com/g...-ohms-law.html
Someone's with more knowledge and practice opinion would be highly appreciated.

[Chumbucket843]

your cpu will last longer if you undervolt because IC's degrade over time but in a real world scenario you wouldnt keep a core i7 for over 10 years because it would be obsolete.

[In-Fluence]

Undervolted several chips without issue, including my E6300 which happily chugged along on 2.8GHz with minimum voltage setting the board would allow (1v), which ended up being my HTPC for a few years. Although I've never experienced problems, I have read that the potential difference between vtt and vcore could cause damage, but that was a long time ago and I never heard any proof.