Do CPUs Fail From Old Age?

[Un4tural]

Hello,

No my cpu isn't failing.

 

I was just curious, as i still got a pentium 3 laying about that still works perfectly, has anyone had a cpu die of actual old age? my E6300 core 2 duo that has been used for gaming overclocked and abused ON A STOCK COOLER is still going perfectly well, stress tested it and everything. (recently cleaned it up, case looked similar to

slightly worse, it's fine after so much abuse, don't ask it has been left for my sister after i upgraded)... the cpu is 7 years old or so, as i got it shortly after release? I've seen plenty of video cards fail, old age overheat etc. never seen a CPU die... To me they seem immortal at this point, considering i can still have a game of red alert 2 on my pentium 3!

 

i am very curious as to what the symptoms of a failing cpu are, does it just go poof and refuse to work or it keeps freezing the system? have managed to extend gpu life time of downclocking it (it became unstable at what were the stock speeds) died eventually, but would that work for a cpu as well? 

[samcool55]

A cpu no, a gpu yes.

Well not sure if it was the gpu itself but a condensator thingy blew up and wrecked the gpu and motherboard. Could be the gpu being putting too much stress on it.

Most of the time the motherboard dies before the cpu. Or the cpu gets so worthless it's thrown away.

[Guest]

I have experienced a CPU dying after about 5 years in my HTPC, it was a core 2 quad that was fairly over clocked and comfortably over-volted (a fair bit but not exceptionally so) but if your temperatures and voltages are within the stated specification then I see no reason for a CPU to die and if it does it would be so old it would be useless. 

[harrynowl]

Not of old age no.

 

Of excessive volts? Yes.

[Un4tural]

i thought i just never saw one by chance, but might they actually be immortal? i got a rubbish sempron 145 which just happens to fit in same socket might just mess with that when  i got time, though it overclocks super poorly, only got around 7% in clock speed before it wouldn't even boot...

[patrickjp93]

They can if overclocked, and in the very very long term they will anyway. Silicon is an unstable atom. Eventually the pathways will fracture due to quantum effects on their own.

[Judist]

Any material you pass electrons through will have a rate of decay. Same with any material that's heated. CPUs get a lot of both.

Chips die from that decay and/or power spikes from unclean electricity, dying or defective vrms, lightning/static, bad power supply, etc.

[Un4tural]

Any material you pass electrons through will have a rate of decay. Same with any material that's heated. CPUs get a lot of both.

Chips die from that decay and/or power spikes from unclean electricity, dying or defective vrms, lightning/static, bad power supply, etc.

 

i understand this, in theory, in practice i never seen this happen to cpu... overclocking and overvolting is understandable, especially over volting...

 

mostly curious of how cpus fail as i never encountered such a thing, i get that in theory it will fail, but in practice it doesn't seem to happen, so i'm just curious.

[Faceman]

I have an AMD 3200+ from I don't even remember what year that is still going strong.  Never overclocked or anything, stock cooler.

[mr moose]

i understand this, in theory, in practice i never seen this happen to cpu... overclocking and overvolting is understandable, especially over volting...

 

mostly curious of how cpus fail as i never encountered such a thing, i get that in theory it will fail, but in practice it doesn't seem to happen, so i'm just curious.

 

I have yet to see any properly design solid state device die of it's own accord.  I have been involved with electronics in one form or another since the 80's and to date there has always been an external influence that has caused failure.

 

E.I if a cpu or any SS device is operating within design spec it should last for at least 30-40 years and then some.

[Un4tural]

I have yet to see any properly design solid state device die of it's own accord.  I have been involved with electronics in one form or another since the 80's and to date there has always been an external influence that has caused failure.

 

E.I if a cpu or any SS device is operating within design spec it should last for at least 30-40 years and then some.

 

 

to be fair i've seen gpus fail, quite a few times, artefacts, flicker etc. never seen cpu die... even after years of abuse dust and running with limited cooling. same with ram, i had a bad stick once, but pretty sure it's just faulty from factory... got ddr sticks that are still working.

 

i agree about SSDs, i was very sceptical getting one, all the horror stories of them failing, limited write cycles, controller failures etc.(in the earlier days) , after owning now 5 of them, not one has failed, works perfectly fine, and from experience with storage so far i trust my data to an SSD over an HDD 9/10 times... exclusions being NAS or long term stoarage, though it's mostly due to cost for NAS. my ssds get plenty of write cycles too... not a hiccup, other than a few BSODs after upgrading firmware on an oldish patriot inferno until it settled. great damn things.