Can you wear out a CPU?

[DoNoHarm]

i think this thread could benefit from higher fan speeds....

 

[m85476585]

You guys are wearing out my network card.

Here's an analogy: Say you have a piece of ice just below freezing. As long as it is below freezing, it will never melt, no matter what temperature it is at (assume pressure is constant, atmospheric pressure). As soon as you let the ice get just a bit warmer, it will melt and it will no longer be ice. Similarly, the damage caused to a CPU at extreme temperatures (above normal operating conditions) is not the same as any degradation that may happen under normal conditions.

 

[dukebound85]

Agree with you? You mean agree with what I originally said.

no, you said running at 100% damages a cpu. it doesnt

Haha wow someone's full of themselves. I've been talking about extreme temperatures being dangerous to a CPU's life and you've been leading this discussion on a wild goose chase full of BS by implying things I never said nor meant. But go on, if that's how you justify your argument.

no, you said running at 100% damages a cpu.......i am not denying high temperatures above the design damage a cpu. 100% is not above the design. You need to understand this basic concept

Gee, I don't know, maybe here:

And you need to understand when I use terms "such as a fan" like a fan, etc that it does NOT equate to me saying a fan is the end all be all of a cooling system...

And if you understand computers at all you'd realize that ANYTHING can fail, ESPECIALLY under extreme heat.

annnnnd running at 100% does not result in extreme heat....

Sorry, reading and trying to understand your maze of implications and fallacies can get tiresome.

Right. If you could simply undertsand that 100% is part of the design of a chip and it doesn't hurt it, we could make some real progress. Why you can't undertsand the csimple concept of design parameters is beyond me

Which is EXACTLY what I said in my first post with this:

tl;dr version: No, just be sure your CPU is getting good ventilation/airflow and keep it as cool as possible (wether that means getting a laptop cooling pad, cleaning out the CPU fans, keeping the room cold, etc.).

except you do not need a cooling pad (pure marketing) though it does help to keep the fans in optimum condition by keeping obstruction (such as dust out of the way)

 

[Stridder44]

no, you said running at 100% damages a cpu. it doesnt

Hahaha wow no I didn't, there you go putting words in my mouth again. As I've been saying this whole time, running a CPU at 100% is perfectly fine. It's the extremely high temperatures that are dangerous.

no, you said running at 100% damages a cpu.......

No, I didn't. YOU said that I said running a CPU at 100% damages it.

i am not denying high temperatures above the design damage a cpu.

Glad that you finally agree with me.

100% is not above the design. You need to understand this basic concept

Please point out where I said "100% is above the design".

And you need to understand when I use terms "such as a fan" like a fan, etc that it does NOT equate to me saying a fan is the end all be all of a cooling system

Ok, a heat sink? Because other than liquid cooling (which is damn near unheard of in a laptop) and just plain old air there isn't much else in the way of components that can cool it off.

annnnnd running at 100% does not result in extreme heat....

No- but, again, I never said it did. But when in a tiny enclosed space for a long amount of time with all the other components equally as hot, which in turn only become hotter, CAN result in extreme heat. No tiny little laptop fan in the world can fix that.

Right. If you could simply undertsand that 100% is part of the design of a chip and it doesn't hurt it, we could make some real progress. Why you can't undertsand the csimple concept of design parameters is beyond me

And there you go yet again! You keep brining up this whole "100% is part of the design" bit, and I don't know why because I said it wasn't in the first place. Why you can't understand that simple point is beyond me.

except you do not need a cooling pad (pure marketing) though it does help to keep the fans in optimum condition by keeping obstruction (such as dust out of the way)

The cooling pad was a simple suggestion, I never said it was necessary (as it often times isn't).

 

[lixuelai]

Video cards don't have moving parts either (aside from the fan) and yet it happens there too. Also I never said anything about a CPU dying from normal use or moderate overclocking. And overclocking is 99% of the time in desktops, and we're talking about laptops. Also, if you had read any of my posts, I'm talking DECADES, not just overnight.

Except it is rarely the GPU itself that failed. Only when there were manufacturing problems like Nvidia's screwup. Also video card is a card. There are other components on it that fail way earlier than the GPU itself does...just like a CPU. Also I did read your posts. First you said that running for days will damage it. Then running for months. Then decades...

 

[rowsdower]

What about electromigration and failure because of accelerated diffusion caused by increased temperatures?

 

[WindowsTraitor]

Look at the intel website

for my processor

http://ark.intel.com/Product.aspx?id=33918

Tjunction: 105degrees C.

On the Data Sheet you will find that there can be "permanent silicon damage due to thermal runway. Therefore you need a protection circuit to shut down the cpu at 124 degrees celcius. (prob measured at tjunction)

http://download.intel.com/design/mobile/datashts/31891401.pdf

In any case by reading the datasheet you will see in the thermal section that the most important is running the chip within the recommended temperatures.

In normal use if you keep the temperatures within those parameters, you will not see the cpu fail in your life time, unless there is a manufacturers defect.(bad batch from the wafer, but the quality control is very high.)

Read also section 5.1.2 about thermal control circuit.

If you want more information about the failure within normal temperature ranges, you can contact their customer service and ask for the MTBF. Mean time between failure.

http://supportmail.intel.com/scripts-emf/welcome.aspx?id=24

I do not know if they will answer you, but you may try.

Nevertheless, if you keep the machine cool, you will not damage it. If you do not keep it cool enough the protections systems, will give you some help, but you may eventually cook your laptop.

 

[m85476585]

According to this site
http://www.eng-tips.com/viewthread.cfm?qid=35348
Intel's goal is <1% failure after 7 years and <3% failure after 10 years. They reference Intel's site, but the link is broken and I can't find that info in Archive.org. It sounds about right, though.

 

[dukebound85]

Stridder Quotes
I am going to quote what you wrote ok
Exhibit #1

There is a high chance that if you're using it at 100% for DAYS at a time, like running Folding@home in a laptop, the laptop could overheat and your CPU would get even hotter, potentially (and most likely) causing damage.

Exhibit #2

I'm saying in a tiny enclosed area like a laptop it's not a great idea to run a CPU at max, 24/7, for days or months at a time (like running Folding@home for example) simply for temperature concerns. I've played games with both CPU and GPU fans running full bore for hours before without a problem whatsoever, but I wouldn't want them doing it for days.

Exhibit #3

I'm just saying running it 100% for a very, very long time will shorten it's lifespan.

Exhibit #4

So a CPU running that is enclosed in an area that is less than 1" with a bunch of other very hot electronic parts at extremely high temps for its' entire life will last just as long as a CPU under the same circumstances running at much lower temps? Wishful thinking.

Hahaha wow no I didn't, there you go putting words in my mouth again. As I've been saying this whole time, running a CPU at 100% is perfectly fine. It's the extremely high temperatures that are dangerous.

See all the above quotes. Those are your words, not mine. I am not putting any words at all in your mouth. You simply forget what you had wrote apparently

No, I didn't. YOU said that I said running a CPU at 100% damages it.

The above quotes, your words, state that running at 100% at extended times damages the cpu. It doesn't

Glad that you finally agree with me.

I am not arguing this. You are stating that running at 100% will cause high temps that will damage the cpu

It won't. Why? Because it is designed for that load. That is why
Case in point: laptops use lower wattage chips, graphics cards, etc compared to dektops

As desktops have more room for cooling capability, they can use more powerful components.

A laptop using its components at 100% will never generate temperature that would damage the cpu until a component like a fan dies

A cpu will not "die" when all other components in the computer are doing their job to keep it cool. It does not degrade by high usage...which is what the OP is concerned about. A component critical to keeping the cpu cool will fail long before the cpu fails from being run at 100%

Please point out where I said "100% is above the design".

Your quotes above state that you belive running at 100% will shorten the life of the cpu if ran on days on end. Your reason is high temperatures from other components and tight spaces, etc

Now I ask you....do you not think that the engineers who designed the computer, picked the components did not analyze the effects other components had on other components?

I assure you that in their analyses, that the other parts and the space confines were incorporated into the design and are adequate for cooling when the cpu is ran at 100% for as long as you want it at 100%

Ok, a heat sink? Because other than liquid cooling (which is damn near unheard of in a laptop) and just plain old air there isn't much else in the way of components that can cool it off.

And you would be wrong

Factors that go into cooling
1) wattage generated by parts and the location of where the wattage is generated in relation to other parts
2) what exact heat sink is needed
3) what size fan? what power of fan? how many fans? fan placement?
4) ventilation? air flow from fans? how to exhaust the air, where are the exhausts headed?

That's just off teh top of my head. It's a little more deatiled then you think it is

This is why I keep asking if you have any idea what design entails and its pretty clear you don't

No- but, again, I never said it did. But when in a tiny enclosed space for a long amount of time with all the other components equally as hot, which in turn only become hotter, CAN result in extreme heat. No tiny little laptop fan in the world can fix that.

Yes you did say that. You said running at 100% for extended periods of time will cause excessive temps. See my posts at the beginning. Your words.

And we covered the second part of your point about the tiny space and fan being designed with 100% usage in mind.

And there you go yet again! You keep brining up this whole "100% is part of the design" bit, and I don't know why because I said it wasn't in the first place. Why you can't understand that simple point is beyond me.

You keep saying that 100% for long times degrades the cpu by creating too high temps. It doesn't as its incorporated in the design. Not a hard concept though it appears it is to some...

The cooling pad was a simple suggestion, I never said it was necessary (as it often times isn't).

Ok, I am just saying it is not needed if you want to run your laptop for 100% for long times.

Except it is rarely the GPU itself that failed. Only when there were manufacturing problems like Nvidia's screwup. Also video card is a card. There are other components on it that fail way earlier than the GPU itself does...just like a CPU. Also I did read your posts. First you said that running for days will damage it. Then running for months. Then decades...

I read them too. Not very consistent to say the least

 

[DoNoHarm]

people; i think that the newer MBP's have safety systems built in in addition to the fans that prevent overheating that damages the CPU. For example, try putting out MBP under a blanket and encoding video. I think you'll find that the temperatures don't deviate from the normal temps. There's clearly an optimal temperature that the apple engineers decided not to exceed!

 

[milton.sheaf]

I am wondering - do CPUs get worn out? I am using my Macbook Pro now as my media server/encoding machine - and it is pretty much always encoding some sort of video and has its CPU maxed out - does anyone know if the CPUs get worn out and slow down over time?

Thanks!

(It's a 17" Intel Core Duo 2.16 GHZ)

In theory yes, but even at 100% utilization, it will be many many years before this happens. I run Folding@Home on my PS3 and Mac Mini and they're on 24/7.

The typical computer has a few wear items that only last a few years, power supplies and fans, PRAM batteries, hard drives. These items have a far shorter lifespan than a CPU. In fact, I've never seen a CPU fail before. In any kind of machine. I have an SGI Indy from 1993 that I have running 24/7 as well as a DEC Alpha from 2001. Both of these work as well as they day they were built. Remember that a modern CPU is a "solid state" device, so there are no moving parts to wear out or burn out.