Sign in with
Sign up | Sign in
Your question

How many years of a computer's life does overclocking kill?

Last response: in CPUs
Share
June 27, 2006 12:36:28 PM

How many years of a computer's life does overclocking kill? I heard someone say it kills about 5 years of a computer's life of 10. Another person said only a week. Of course, I presume that there is a good water cooling kit and fan setup in the computer. If you have a good link, pleae post. :D 
June 27, 2006 12:50:14 PM

i dont hav a link but it would depend on how much u overclock it/how hot it runs.
every cpu will be different but the average cpu life is about 10 years(i think)(it might be 15) so it might take 3-5 years maybe.
i dont know much about overclockin though
June 27, 2006 12:59:18 PM

There is no answer, but if you have to ask then you shouldn't do it.

Many things can effect it, most notibly how much voltage is used.
Related resources
June 27, 2006 1:07:22 PM

Intel quote a failure rate of 1 in 100 during the first 10 years of life, after that they give no guarantee / advice... I guess the damage you do to a CPU is more related to how hot you let it get, and a little bit of random luck...
Anonymous
a b à CPUs
a b K Overclocking
June 27, 2006 1:56:39 PM

As stated, I think its primarly a heat issue and them comes the fact that the gates are switching faster than they ought to.

My northwood has been oced for close to 4 years now, but its running at 35-40 at full load, hardly hot, so I dont expect to have it die on me. On the other hand my 865 motherboard is starting to feel old, I think the northbridge is getting tired of operating at this speed...

Bottom line is if you OC without raising the vCore or at least not raising it too much, many other parts should die before the CPU and your CPu will probably be obsolote by the time it dies anyway!
June 27, 2006 2:32:15 PM

Quote:
How many years of a computer's life does overclocking kill? I heard someone say it kills about 5 years of a computer's life of 10. Another person said only a week. Of course, I presume that there is a good water cooling kit and fan setup in the computer. If you have a good link, pleae post. :D 



if you run it out of spec (heat) it will shorten the life. if you keep it cool it shouldn't be aproblem.
June 27, 2006 2:38:34 PM

overclocking an amd 64 0.2Ghz without changing the default voltage and causing a temp difference of only 1-2C won't shorten a minute of it's life.
June 27, 2006 2:39:50 PM

In reality, as long as it doesent get too (definitions of too depending on the particular max thermal limits of the cpu in question) hot, its life wont be shortened enough for it to matter, since most ppl are unlikely to keep the same machine long enough for the processor to fail, even if its heavily overclocked, and if it is heavily overclocked people would likely have some bizarre cooling system anyway, and the kind of people that do that are the kind of people whoe frequently buy new computers, which further proves the point that it doesnt matter.
June 27, 2006 2:52:36 PM

Read the old, yet still informational, THG article on overclocking.. Or, just search the web for 'electro migration', which will occur with or without overclocking, but is hastened by increasing the speed and/or voltage on the chip. It's not necessarily heat, though that will also speed up the process.
a c 448 à CPUs
June 27, 2006 3:26:35 PM

As long as you do not do extreme overclocking and as long as you keep the CPU relatively cool, it should be fine. Chances are that you will upgrade before the CPU fails People usually do a major upgrade every 4 years or so.
June 27, 2006 4:55:53 PM

Chances are that 99% of people who overclock (i.e. enthusiasts) are going to buy a new system before their crimes against CPUs have caught up with them
June 27, 2006 5:16:08 PM

Quote:
How many years of a computer's life does overclocking kill? I heard someone say it kills about 5 years of a computer's life of 10. Another person said only a week. Of course, I presume that there is a good water cooling kit and fan setup in the computer. If you have a good link, pleae post. :D 


where the f*uck do u get your information? high school kids?
June 27, 2006 6:36:16 PM

I doubt if anyone here has kept the same computer for 10 years. If you kept one for 10 years what would run on it now ?
NOTHING
June 27, 2006 6:46:38 PM

Can you changing the clock speed anytime you have the computer, or, say you buy a computer and later on getting a cooling system, can you then clock it again?
June 27, 2006 6:49:46 PM

Well... generally speaking you cannot overclock OEM systems. However, there are free utilities out there than can change the bus speed within windows.

So the answer is no. But only if you have an OEM system. Custom built & self built systems can generally be overclocked at any time.
June 27, 2006 7:19:00 PM

im using a pentium one laptop right now. i just use internet and listen to music with aim firefox and winamp. ehehhehehehe
i also am using a dlink wireless card. this thing is faster than a lot of peoples spyware infected computers. 166mhz 16mb ram w00t pwns
June 27, 2006 7:21:03 PM

Quote:
There is no answer, but if you have to ask then you shouldn't do it.


A lot of people say that but I don't really get why. If you're going into overclocking you would want to get a vague idea on what happens.

A couple here have said it depends purely on heat, but I'm pretty sure, even running the cpu at 0C wouldnt extend the life if you were running it at 1.65v.

I found a detailed thread about it somewhere on another forum - I'll try and find that but I'm pretty sure it concluded:

Due to electromigration, voltage shortens a processers life more than Heat which shortens a processers life more than frequency.

Maybe in the coming years, someone should try and construct a database of how long cpus take to die, or how often, if they have been overclocked carefully (that doesn't mean putting high voltages thorough the cpu, it means not making mistakes.)

I'm not sure I've read of any, 'my cpu just randomly died', with the answer being 'thats the risk of overclocking'.

Its always a psu malfunction, or dodgy thermal paste. ..or something

A lot of what Ive said is probably misguided. . .but my 2 cent anyway :) 
June 27, 2006 7:22:55 PM

Quote:
How many years of a computer's life does overclocking kill? I heard someone say it kills about 5 years of a computer's life of 10. Another person said only a week. Of course, I presume that there is a good water cooling kit and fan setup in the computer. If you have a good link, pleae post. :D 


where the f*uck do u get your information? high school kids?

:lol:  . o O (Can someone please tell me where the Pet Cemetary is? My CPU died and I thought I'd try to bring it back to life, and yes... I know I need to bury my own)
June 27, 2006 7:28:13 PM

Quote:
There is no answer, but if you have to ask then you shouldn't do it.


A lot of people say that but I don't really get why. If you're going into overclocking you would want to get a vague idea on what happens.



It's my personal opinion that one could get more than a vague idea of what might happen by simply reading and searching the internet, and therefore I do think often times people need know more about things technically than what they might gleen from a few posts on an internet forum. That's not to say to not read internet forums, but don't accept everything posted as the truth, including what I post, cause I lie. :wink:

Quote:
I'm not sure I've read of any, 'my cpu just randomly died', with the answer being 'thats the risk of overclocking'.

Its always a psu malfunction, or dodgy thermal paste. ..or something


Exactly why I think much of what one might read on an internet forum should be taken with a huge grain of salt, because a lot of it is from those parroting back wrong information they've heard from a friend of a friend.
June 27, 2006 7:35:36 PM

Quote:
because a lot of it is from those parroting back wrong information they've heard from a friend of a friend.


More likely, information that someone just made up. ;-)
June 27, 2006 10:13:21 PM

I stand corrected, there is stuff you can run, but one man's laptop is another man's paperweight lol. I'd get tired of waiting for it to boot.
June 28, 2006 12:19:00 AM

Why do you have 2 power systems in your rig?
June 28, 2006 11:58:27 AM

The 145 Watt runs the fans on my water cooler radiator, hard drive, power regulator system. My cooling system is independent of the power system for the computer, the water and cooling system powers on first, then 5 seconds later the computer boots. That way I can make sure that there is circulation before power up.
My video card has a chamber built around it also to cool the memory, the GPU is watercooled but the memory isn't. I push air into it, then another fan pulls it out across the memory heatsink and regulator circut. I can run the video memory up over 1200 MHz, but the returns are minimal at best.
All the bugs are out of this system for over a year and I will be transferrering this cooling over to my new system when I decide on what hardware I'm going to use.
June 28, 2006 1:13:12 PM

it`s not just the temperature my friend .... it simply forces the original design to work at same other level......higher voltages are applied etc......

no matter how you put it overclocking, whether it shortens the life or not, it`s an out of warranty issue, it`s unproper used device......

good luck
June 28, 2006 8:55:10 PM

Quote:
I doubt if anyone here has kept the same computer for 10 years. If you kept one for 10 years what would run on it now ?
NOTHING


so true
June 29, 2006 12:20:12 PM

Not true

I have just upgraded from AMD K6 266 this year. And guess what you could get on the internet with it too! :lol: 

It also ran Win 2k (rather slow but it worked)

And with a 3DFX graphics card it played games.

OK it was only 9 years old but not far off :wink:
a b à CPUs
June 29, 2006 1:26:30 PM

If your reasonable with the your expectations, how far you OC, and ensure proper cooling; epsecially since your asking how many years it would take to kill the proc, I don't understand why there is any concern over a reduction in proc lifespan...hell, I had an oc'd Skt 423 from 1500MHz to 1850MHz with a self modified heatsink and fan setup and I ran that for 2+ years without any issues, squeaks, burps, or farts...

But as noted...
Quote:
...overclocking...it`s an out of warranty issue, it`s unproper used device...

And we wouldn't want to be "unproper", now would we...

Good luck!
June 29, 2006 1:48:00 PM

Law of Physics – Thermal Breakdown

OHMS Law
More frequency means more voltage ----- Hertz
More voltage means more current ----- Volts
More current means more power ----- Amps
More power means more heat ----- Watts


How much energy can you put on a 24 AWG 'gauge' of wire before it burn/break/melt?

CPU has 1000 time smaller of cross connection of wires.
June 29, 2006 1:52:20 PM

Quote:
How many years of a computer's life does overclocking kill? I heard someone say it kills about 5 years of a computer's life of 10. Another person said only a week. Of course, I presume that there is a good water cooling kit and fan setup in the computer. If you have a good link, pleae post. :D 


it's not the actual OC'ing that shortens the life of a cpu but the being precise, cautios and having a bit of luck with the parts you get... i once killed a new cpu in it's first few hours of operation (forgot to put the HF on the cpu and that was years back where it wasn't standard having temp-sensors everywhere ... :x )


i had once a 80486 @ 10 MHZ which u could OC to 33MHZ with a button on the case. the system came with neither heatsink nor fan on the CPU and the system only ran about 5 minutes stable @ 33MHZ before it crashed.

that's why i used a bare ice-pack (usually used for injuries) to put directly on the cpu.... this way that system ran OC'ed (330%) for about 2 years before some misterious water-drops on the socket caused a short circuit :wink:
June 29, 2006 2:26:19 PM

:D  chyba jestes smiesny/??
It only kills if you amp up the voltage, and if it gets too hot, other than that getting good coolage for your system is a nesessity for and OC rig.
June 29, 2006 2:41:38 PM

Quote:
Law of Physics – Thermal Breakdown

OHMS Law
More frequency means more voltage ----- Hertz
More voltage means more current ----- Volts
More current means more power ----- Amps
More power means more heat ----- Watts


How much energy can you put on a 24 AWG 'gauge' of wire before it burn/break/melt?

CPU has 1000 time smaller of cross connection of wires.


What does Ohms law have to do with frequency?

Quote:
i had once a 80486 @ 10 MHZ which u could OC to 33MHZ with a button on the case. the system came with neither heatsink nor fan on the CPU and the system only ran about 5 minutes stable @ 33MHZ before it crashed.

that's why i used a bare ice-pack (usually used for injuries) to put directly on the cpu.... this way that system ran OC'ed (330%) for about 2 years before some misterious water-drops on the socket caused a short circuit :wink:


Do you mean the Turbo button? If I recall, the turbo button was used to slow the machine down...underclock it if you will. You had a 33Mhz 486 that you could slow down to 10Mhz, not the other way around. :wink:
June 29, 2006 3:24:06 PM

well even if I had to underclock the cpu with the button it was still a really strange machine because it just wouldn't run stable, at least not until i'd put the ice-pack on the cpu...

but still.... the good old times where it was sufficient to put something cold on the cpu to cool it down :roll:
June 29, 2006 4:07:59 PM

Also keep in mind that not just overclocking can kill a CPU. Thermal expansion also plays a part. During a computers life span of being turned on and off X amount of times, you may cause damage. Electonics expand ever so slightly with heat from being powered on, and shrink back ever so slightly after being shut off and and left to cool. This back and forth will stress all parts not just the CPU over time. After working in a electronics lab for a few years and doing thermal testing...CPUs can be heated quite hot and still function fine. Its the small things like ceramic resistors and capacitors that change values with heat that can cause problems and intern "kill" a CPU.
June 29, 2006 5:56:23 PM

Quote:
Do you mean the Turbo button? If I recall, the turbo button was used to slow the machine down...underclock it if you will. You had a 33Mhz 486 that you could slow down to 10Mhz, not the other way around.


Yes I remember it that way as well, the turbo button throttling your CPU so that your 486-33mhz didn't run too fast and crash your program!
July 9, 2006 6:38:08 PM

Ohm's Law is the necessity of making measurements of voltage and current that are averaged over a sufficiently long period of time in order to compute accurately a resistor's actual resistance. This constraint would not have been understood by Ohm in the 1820s because instruments for making current and voltage measurements over short periods of time were not available. A hundred years later, the effect now known as Johnson noise, or thermal noise, was discovered. This thermal effect implies that measurements of current and voltage that are taken over sufficiently short periods of time will yield, usually, ratios of V/I that are different from the time averaged value of R, although the ensemble average of the ratios computed from a large number of statistically independent current and voltage measurements will yield the time-averaged value of R.

Ohm's work long preceded Maxwell's equations and any understanding of frequency dependent effects in AC circuits. Modern developments in electromagnetic theory and circuit theory do not contradict Ohm's law when they are evaluated within the appropriate limits.

In a sense, Ohm's law is not a law at all. Rather, Ohm's law may be more properly seen as the definition of resistance. (In the hierarchy of 'physical' or 'logical' terms, definitions are more fundamental than laws or theories.) Joules, coulombs and time are basic units of measure. A volt can be defined based on those units of measure. Ohm's "law" defines the ohm. An ohm, the unit of resistance, is derived from those more basic units. In essence, Ohm discovered that V/I is a constant for a given conductor at a given temperature, and over a large range of V and I.
July 9, 2006 7:10:33 PM

zero if you know what you are doing. But it's a stupid question since none can answer it since it is different from system to system.
July 9, 2006 7:24:16 PM

How long a life does my Opteron 175 have at stock? :wink:
July 9, 2006 7:38:23 PM

Quote:
How long a life does my Opteron 175 have at stock? :wink:


now thats not fair. :lol: 

the way you take care of that rig your opty will out live most stock cpus
July 9, 2006 7:43:19 PM

Any 939 Opteron owner should be shamed for running stock, and that shame will destroy the CPU's will to live...
I run mine at 2.64GHz 24/7!
July 9, 2006 7:46:42 PM

I couldnt agree more.

they love to perform.
its like useing a Farari to go to the grocery store.
July 9, 2006 8:01:06 PM

Rexter, thanks for the Ohm's law refs, but I think you'll find that Johnson–Nyquist noise is not normally considered to be part of Ohm's Law (which typically is just listed as V=IR and P=(I^2)R ). In any case, it's not thermal breakdown that's the primary effect under the regimes we're dealing with, but electromigration, as linked to by Pain in message #10.
July 9, 2006 9:11:44 PM

Quote:
I doubt if anyone here has kept the same computer for 10 years. If you kept one for 10 years what would run on it now ?
NOTHING

A P2 Windows 98SE running a CS server1.6 for my son is over 10 years old. My Penium 60Mhz running windows 95 is a storage box and its going on 13 years old.

OCing can burn out a CPU is under 2 years, is the simple answer, due mostly to CPU designs having some structures running higher temps than others were those areas are dissipate thought the CPU. At spec temps OCing increases the already over heating of some structures and in the end causes, burn though, leakage for those areas and finally death for the CPU.

Finding these hot spots is part of what Intel and AMD call the refinement process. Finding those areas and redesigning them for higher clocks are an ongoing process.

OCing a CPU really requires a lower than stock temp to keep high heat areas from critical temps. OCing can reach very high performance without risking much life by once you find a CPU's breaking point with better cooling there is a need of backing off from the max until the CPU is below the stock temps. How much below isn't easy to answer but I try to lower the temp from stock by the same percent of the OC.
July 10, 2006 1:25:20 AM

Indeed it is not, I should have been more specific, even though I wrote it in my first post I stand corrected. My apology to all and I manage to skip that post by Pain, thanks for directing me to it.
!