Settle This

[Elmo]

Does OCing a Cpu reduce the lifespan of the chip? Please post your opinion because ive been hearing alot of mixed comments about this where some people have said that their cpu overclocked after a certain time cant cope with the cpu clock and they are forced to lower it to prevent crashing..

[D007]

Yes... Electron migration due to increased voltages, increased clocks and increased heat.. IMO absolutely it does.. Not that it will matter in the end, especially if you upgrade on any kind of semi-decent basis.. It's not like you buy a cpu that will live 20 years and it will die in 5 now.. It may very well last 19 years.. The rate of degradation varies but it still kills it a bit..
You really have to spceify wether you mean a moderate OC or an extreme OC..
Even an extreme OC if cooled properly, can get "most" of it's original life..
But not all... That's what I gather anyway..

[[Ion]]

Almost certainly. CPUs are designed to run at certain specifications, and running them over this will shorten their life. Higher temperatures and higher voltages certainly cannot help. As long as the overclock is relatively mild, however, this isn't an issue; CPUs can last a very long time (I have a Pentium 3 laptop that still works fine), and almost everyone will upgrade before the CPU would die (particularly enthusiasts, who will be the ones doing the OCing in the first place).

[Elmo]

Well i heard someone with a 3930k at @ 5 ghz on liquid had it fine for a few months until it started to crash so he decreased his clock speed to 4.9 and after sometime the same pattern.. until today he said its 4.6 now.. So i guess you are saying even if i use ln2 or liquid cooling it does not infact help reduce the wear and tear of a chip?

[assaulter_99]

Yes, judging by how quickly afterwards my computer stopped shutting down altogether. Left me without answer as to why. Couple of other problems afterwards! On the bright side though, it still runs fine. I've gave up on trying to find the problem as to why I must reseat all RAM sticks before powering up my pc or else it doesn't even post, since its gonna go in the graveyard shortly! All of this started after approx 12 months into some heavy OC'ing and it was on 24/7. Damn time for some upgrades anyway. And it was all on air... Folks with liquid might get better results.

[brandonwh64]

The higher the voltages the more the degradation I have seen. Keeping a chip at a moderate voltage has only shorten the lifespan a fraction compared to high voltages for benchmarks. This is why I do not buy benching chips for daily drivers if you get what I mean. Find a clock at the voltage you feel comfortable with and stick with it cause more than likely it will last until your next upgrade.

My I7-920 @4ghz 1.275V lasted a long time with no issues.

[Elmo]

Originally Posted by brandonwh64

The higher the voltages the more the degradation I have seen. Keeping a chip at a moderate voltage has only shorten the lifespan a fraction compared to high voltages for benchmarks. This is why I do not buy benching chips for daily drivers if you get what I mean. Find a clock at the voltage you feel comfortable with and stick with it cause more than likely it will last until your next upgrade.

My I7-920 @4ghz 1.275V lasted a long time with no issues.

aight thx well i wasnt sure if i should even oc my future build well i normally upgrade my procs ever 2 years . So i guess it should be fine :S

[Jizzler]

Originally Posted by Elmo

Well i heard someone with a 3930k at @ 5 ghz on liquid had it fine for a few months until it started to crash so he decreased his clock speed to 4.9 and after sometime the same pattern.. until today he said its 4.6 now.. So i guess you are saying even if i use ln2 or liquid cooling it does not infact help reduce the wear and tear of a chip?

- Was the OC obtained during the winter when there is lower ambient temps?

- Regular blow-out of the radiators being done?

- Was the OC thoroughly tested? Testing with multiple synthetic and real applications is key. Sometimes you can go indefinitely at "full load" with one app, but crash under another.

[natr0n]

It depends on the chip and the cooling period.

some oc's last just about forever some don't.

[brandonwh64]

Originally Posted by Elmo

aight thx well i wasnt sure if i should even oc my future build well i normally upgrade my procs ever 2 years . So i guess it should be fine :S

Another thing I would like to add is that I have a 2600K that was purchased at launch that has crunched 100% load at 4.5Ghz 1.33V 24/7 (offline only in outages) and still is rock stable from the day it was purchased.

[dark2099]

One thing to add to this, I cannot myself provide any evidence either way, but have read about it a couple times, constant stress testing/stability test will make the degradation worse. The reason behind that is that you are making the CPU run at its absolute max speed, voltages, and thus make it produce the max heat it can, for an extended period of time. Once again, many variables to take into account to factor how much time is being taken off the life, but these things do make it worse.

[assaulter_99]

Originally Posted by Elmo

Well i heard someone with a 3930k at @ 5 ghz on liquid had it fine for a few months until it started to crash so he decreased his clock speed to 4.9 and after sometime the same pattern.. until today he said its 4.6 now..

Thats my pattern. Started at 3.6 then slowly had to decrease it over time (3.5,3.4,3.2,3.0). Everything will hang if I get over 3.0 now (will boot fine, but as soon as I use some real power, some weird stuff happens). I am now @ stock since I'm not bothered with it anymore.

[Elmo]

Originally Posted by Jizzler

- Was the OC obtained during the winter when there is lower ambient temps?

- Regular blow-out of the radiators being done?

- Was the OC thoroughly tested? Testing with multiple synthetic and real applications is key. Sometimes you can go indefinitely at "full load" with one app, but crash under another.

No

Yes

Yes

[Jizzler]

Well, then is it like the other posters experience (gradual reduction in stable speed).

I wonder if the overclock had started at a lower speed, say 4.8Ghz, would the reduction to 4.6Ghz still occur? Unfortunately I don't have the means like I used to do such an experiment. Typically for my own machines I find a nice comfortable overclocked speed and I'm done. They last 4-5 years at that speed. Could probably go longer, but at that time it's no longer worthwhile to run such an old computer.