Over-clocking/Warranty

[seekermeister]

I know that in the past, over-clocking would void the warranty on many components...is that still the case now?

 

[Airbot]

Why would it have changed? Unless they (manufacturer) specifically state that you can and still be under warranty, I'd assume it would still be the same.

 

[seekermeister]

Partly because many of the components that are made today are specifically designed to be over-clocked, and the manufacturers often tout them for that very fact.

 

[Airbot]

Eh, I'd still say the same applies - Unless they (manufacturer) specifically state that you can and still be under warranty, I'd assume it would still be the same.

 

[seekermeister]

Probably a difficult question to answer, but does any manufacturer so state in their warranties that over-clocking is permitted?

 

[simmo]

it voids warranty...end.

cheers

 

[Fumz]

Probably a difficult question to answer, but does any manufacturer so state in their warranties that over-clocking is permitted?

No. Overclocking still voids the warranty. It's misleading for sure because they'll tout the benefits of overclocking and show you how well cpu's overclock... but... it's all still at your own risk because no two cpu's are identical.. so there are no guarantees.

 

[seekermeister]

The other part of the question is how would they know if a component was over-clocked or not...assuming that it wasn't burned.

 

[sygnus21]

Partly because many of the components that are made today are specifically designed to be over-clocked, and the manufacturers often tout them for that very fact.

That's true, but it's a catch 22 - they design them that way to push sales, but void the warranty if the item failed due to pushing past it's stated design limits - eg we all know you can push say a 2600K safely to 4gig+, even though Intel spec'd it to 3.4. Now tell Intel the item failed at 3.9gig and..... sorry but that chip is not designed to run at that speed so no warranty replacement for you.

Again, marketing - who in the enthusiast community is going to buy a chip that can't be overclocked? And before you say most average users don't care about overclocking, yeah, that's true..... they also aren't replacing components at the pace enthusiasts do.

Is it sneaky, yeah, but these companies aren't specifically telling you to overclock their hardware, they're just not saying it can't be overclocked.

So again, a catch 22

The other part of the question is how would they know if a component was over-clocked or not...assuming that it wasn't burned.

They probably wouldn't unless they suspected it and did a close examination, but.....

 

[seekermeister]

To me, that is like a car manufacturer selling cars with a governor attached to the carb, and touting the speed that the car is capable of...without the governor. Providing special tools to remove or disable it, and never saying a word about the fact that they will void the warranty of the entire auto if you do, even if the failure that later occurs has absolutely nothing to do with the performance tweak.

In real law, if there is no specific law that says that you can't...you can, but in this instance you can't, despite the fact that there is no specific law that says that you can't. It is as though they hope that you will, so that they can void the warranty, and shirk their responsibilities.

 

[sygnus21]

It's a win win for them - they push out an item knowing it can be overclocked so people will snatch them up, but don't have to warrent them if it breaks since they never advertised you could overclock the item in the first place.

You buy & overclock, they don't have to replace it if it breaks - Income in, nothing out

 

[Fumz]

There is a "law" saying you can't overclock. Intel and AMD say if you overclock the cpu, you void the warranty... that's the law.

It's actually pretty reasonable... after all, can you imagine the losses if they were to replace every cpu damaged due to user error and incompetence/ignorance?

 

[sygnus21]

It's actually pretty reasonable... after all, can you imagine the losses if they were to replace every cpu damaged due to user error and incompetence/ignorance?

Yeah they'd go out of business pretty quick on that one

 

[Wishmaster]

Some manufacturers not only use OCing as a sale tactic, but also specifically state its OK to Overclock thier products without voiding your warranty.
Although I think they mean within reason. Intentionally putting double the safe MAX voltage for example may not be covered...

For example, EVGA GPUs and MOBOs.

Question / Issue

Does overclocking void my warranty?
Answer / Solution

Overclocking our products does not void the warranty as long as there is no physical damage to the product or missing components. However EVGA Support will not be able to assist you in overclocking the product.

If you start doing extensive modifications or toss it off a building, give it a shower, etc that will void the warranty though.


Then again, some specifically state OCing will in fact Void the warranty, regardless what advertising tactics they use or are rather vague about it.

 

[seekermeister]

There is a "law" saying you can't overclock. Intel and AMD say if you overclock the cpu, you void the warranty... that's the law.

It's actually pretty reasonable... after all, can you imagine the losses if they were to replace every cpu damaged due to user error and incompetence/ignorance?

If a product is damaged in any fashion, including over-clocking, caused by user error, that is a reasonable basis for denying warranty service, but it is obvious that over-clocking is not necessarily the cause for all defects. Take a processor for example, it could be damaged by over-clocking, but if it were, it would be obvious due to signs of burning or possibly chipping.

I have no problem with the idea of them protecting themselves from their customer's abuse, it's just that over-clocking is not automatically a justified cause for that kind of mindset.

 

[Fumz]

There is a "law" saying you can't overclock. Intel and AMD say if you overclock the cpu, you void the warranty... that's the law.

It's actually pretty reasonable... after all, can you imagine the losses if they were to replace every cpu damaged due to user error and incompetence/ignorance?

If a product is damaged in any fashion, including over-clocking, caused by user error, that is a reasonable basis for denying warranty service, but it is obvious that over-clocking is not necessarily the cause for all defects. Take a processor for example, it could be damaged by over-clocking, but if it were, it would be obvious due to signs of burning or possibly chipping.

I have no problem with the idea of them protecting themselves from their customer's abuse, it's just that over-clocking is not automatically a justified cause for that kind of mindset.

True, overclocking is not the cause of all defective cpu's... sometimes cpu's are defective when they arrive at your door. However, it's untrue that a fried cpu would show obvious signs. Yes, it could show signs, but just as often it could come out of the machine as pristine looking as when it went in.

It's not heat that kills cpus, it's volts. This is true whether you're overclocking or not. Eventually, all cpu's succumb to electromigration.

Intel can only guarantee their parts to operate for x years under a given voltage. Anything outside that voltage and it's anyone's guess as to how long the cpu will operate... and therein lies the risk... you just never know when the volts are going to kill the cpu. It's going to happen; when is the question?

Most guys who OC replace their chips long before they notice the ill effects of pushing too many volts to the core, but, guys kill brand new cpus all the time and there aren't any dramatic pics of fire and smoke, because there's just no outward signs of damage. Again, no two cpu's are alike, so, given two identical cpu's, two identical boards and identical voltage, one chip may run for the next 10 years; another, only for the next 10 weeks.

Given this reality, it's perfectly reasonable for AMD and Intel to have adopted the mindset they have.

 

[seekermeister]

Correct me if I'm wrong, but isn't volts and heat tied together? Yes, I know that a processor can be overheated by poor cooling, even at normal voltage, but if properly cooled, the reason for overheating is voltage adjustment, they are the flip-side of the same coin. Evidence of overheating is obvious even to an amateur, if they know what to look for...a darkening of the PCB on the bottom. I doubt that neither AMD or Intel have crystal balls that permit them a more definitive diagnosis without dismantling a processor, and I doubt that they ever do that for an RMA inspection.

 

[Fumz]

Correct me if I'm wrong, but isn't volts and heat tied together? Yes, I know that a processor can be overheated by poor cooling, even at normal voltage, but if properly cooled, the reason for overheating is voltage adjustment, they are the flip-side of the same coin. Evidence of overheating is obvious even to an amateur, if they know what to look for...a darkening of the PCB on the bottom. I doubt that neither AMD or Intel have crystal balls that permit them a more definitive diagnosis without dismantling a processor, and I doubt that they ever do that for an RMA inspection.

Again, it's not heat, per se, that kills a cpu.

Heat is a function of voltage and clock speed. Yes, you can increase temps by increasing voltage, but you can also increase temps by increasing clock speed without touching voltage.

Cpu's shut themselves off when they reach a certain temperature, so it's really very rare to see a cpu that was killed by heat. For all sorts of reasons you want your cpu to run as cool as possible, but heat alone won't kill a cpu.

One cannot tie heat and voltage together here. Yes, they're related, but really, it's all about the volts. If you were to look at a years worth of dead cpus, I'll bet my left arm only a small fraction of them would be the blackened scarred type you see on youtube... complete with the firemarshall in the background. Fact of the matter is most times the only way you can "see" that a cpu is dead is indirectly; because the machine won't turn on.

 

[seekermeister]

Okay, if a processor has been killed by either volts or clock speed, without causing overheating, how would the manufacturers know that upon inspection?

 

[Fumz]

Clock speed alone won't kill a cpu. Clock speed would only generate heat; enough heat and the cpu would either throttle down or shut itself off.

There are some pretty sharp cookies over at Intel. I have no idea what they do, but if they wanted to know what killed a cpu, my guess is they would find out.

lol, you gotta drop the notion that heat is a culprit.