Does Voltage Kill CPUs?
- Thread starter Greyhound
- Start date
Yes it supposidly does but i doubt you will want the cpu in 8 years time anyway.
Have a search on the forums for electro migration.
Sorry not sure about voltages on the opty but i wouldnt like to put more than 1.5V through my 4400 on water. Although 1.55 would be ok i Guess.
Have a search on the forums for electro migration.
Sorry not sure about voltages on the opty but i wouldnt like to put more than 1.5V through my 4400 on water. Although 1.55 would be ok i Guess.
Soldato
I've got 1.66v running through my X2 3800+ on water, if it breaks i'll just buy a new one. Its not something I worry about 
Just the heat!Greyhound said:
I don't think that is correct. The heat will get it everytime, that is why manufacturers publish thermal guidelines. . however I do agree that overclocking will probably shorten the lifespan of said CPU (live fast die young sorta thing!) but this is a moot point as I don't think anyone will use a CPU for more than five years.welshtom said:
I'm sure a reply will come from a guy still running an overclocked Celeron 300a lol
Soldato
Big.Wayne said:
It is, too much voltage even when running at 30c can kill a chip. You're straining the silicon.
Don
Heat and/or voltage reduce the lifespan of a CPU. Intel/AMD design their chips to last say 15 years within a set of heat and voltage tolerances, say 60C and 1.4V for a A64. Go over either for prolonged periods and you reduce the lifespan.
Jokester
Jokester
Has to be said. Even the guide that this forum pimps to newbies doesn't mention electro migration.
Yeah, there's a suggestion that too much voltage is bad, but nothing's quantified clearly. IMO this is especially dangerous in the hands of a newbie (like me and many others).
Call me Devil's Advocate, but perhaps it's time for one of the board 'elders' to write a unified theory of overclocking guide?
3oh6 said:
Yeah, there's a suggestion that too much voltage is bad, but nothing's quantified clearly. IMO this is especially dangerous in the hands of a newbie (like me and many others).
Call me Devil's Advocate, but perhaps it's time for one of the board 'elders' to write a unified theory of overclocking guide?
Associate
welshtom said:
I've seen the Germans put 2.2v through a water cooled A64, BEAT THAT!!
heat & voltage go hand in hand,
think of it like this, the hotter a conductor, the more resistive it is, the more resistive, the less current can pass through before you get electromigration problems (basically exceeding the current density of the conductor).
Electromigration basically can break down the conductor (effectively etches it over time).
More voltage is required to allow the same currrent to conduct due to the higher restistance, this is why upping the vcore helps overcloking stability.
now if you can keep the heat down, you keep the resistance down, this is why things like phase change cooling allow higher overclocks, resistance has dropped therefore the current flowing in the conductors does not get near the density to cause electromigration problems, even at higher speeds (which generate more heat as the semiconductor switches are operating faster therefore generating more heat).
think of it like this, the hotter a conductor, the more resistive it is, the more resistive, the less current can pass through before you get electromigration problems (basically exceeding the current density of the conductor).
Electromigration basically can break down the conductor (effectively etches it over time).
More voltage is required to allow the same currrent to conduct due to the higher restistance, this is why upping the vcore helps overcloking stability.
now if you can keep the heat down, you keep the resistance down, this is why things like phase change cooling allow higher overclocks, resistance has dropped therefore the current flowing in the conductors does not get near the density to cause electromigration problems, even at higher speeds (which generate more heat as the semiconductor switches are operating faster therefore generating more heat).
Associate
Mr Mister said:
Especially with the looks of the new intels
No swearing on the forums please
Off topic, ive been invited out in Ilkeston next weekend, might go to Notts instead.kimandsally said:Especially with the looks of the new intels
No swearing on the forums please
Who would keep their CPU for 15 years anyway, personnaly i would rather have better performance for 2 years.
Soldato
- Joined
- 9 May 2005
- Posts
- 2,511
pegasus1 said:
Quoted for the truth, I have run up to 1.75v through processors on good air 24/7 and havent had one fail yet
Soldato
mcmad said:
And it's important to remember that the whole CPU isn't uniform. Imperfections mean some areas can suffer enhanced "etch" for years without a problem, while others will break down more quickly. If those weak areas are in a critical place (I'm not sure how much of a CPU core *isn't* absolutely critical!
then an overheated/overvolted CPU might die very quickly. It's a luck of the draw thing.Andrew McP (whose Celeron 300a is still running@450 quite happily in his mother's PC. She doesn't exactly push it hard though.)