Overclocking lessens a CPU's life span, but whos ever had a CPU die on them before?

[Maximilian]

Seriously? Has anyone actually ever had a CPU die before just through simply using it normally, i.e not overclocking or anything? If so, why would it die? What is it that wears out and stops it working?

 

[RichUK]

think of the main variables and that is you answer

 

[almach1]

it'll outlive it's usefullness and be outdated before it goes bad

 

[Maximilian]

Yes but why does it "go bad"? I know they outlive their usefulness, there are still 486's and older running out there. When a bulb dies, its the string thingy inside that burns out, so what it is that dies in the CPU? Could it be any number of things? Or is there 1 thing inside that will simply stop working after time, like the cache, or a certain register or w/e?

 

[Hacp]

I think its something like the increased voltage deteriorates the transitors over time or something along those lines. I have no knowledge of this, but I heard it from somewhere .

 

[BigCoolJesus]

Yea, the transistors wear down over a period of time, and then break


think of it like this, a transistor on a cpu is ungodly small, so the "legs" are even smaller (like a lightbulb filiment, but scaled down even more)
So, just like a lightbulb, the "legs" will eventually wear down from the voltage and break, severing the connections.

That is why overclocking hinders lifespan, because more voltage is being pushed through those already stressed out "legs" (and im sure if enough voltage was used, the transistor itself could burn out)

 

[WobbleWobble]

I've seen a Celeron 400A die off, never been o/c'd or nothing.

 

[Killmenow]

I have seen 2 CPU's die........ *tear tear*

 

[Maximilian]

Originally posted by: BigCoolJesus
Yea, the transistors wear down over a period of time, and then break


think of it like this, a transistor on a cpu is ungodly small, so the "legs" are even smaller (like a lightbulb filiment, but scaled down even more)
So, just like a lightbulb, the "legs" will eventually wear down from the voltage and break, severing the connections.

That is why overclocking hinders lifespan, because more voltage is being pushed through those already stressed out "legs" (and im sure if enough voltage was used, the transistor itself could burn out)

I see. I assume these "legs" are made of copper? Wonder if Intel could custom make one cpu where they are made from titanium! or gold! and it would never die. That would be pretty sweet.

 

[theMan]

i had a cpu die once. it was a celeron 300mhz. it just died. no overclocking. defective. but it was under warrantee. otherwise, they work for years. 10, 20 years. with or without overclocking.

 

[RichUK]

Originally posted by: Soviet

Originally posted by: BigCoolJesus
Yea, the transistors wear down over a period of time, and then break


think of it like this, a transistor on a cpu is ungodly small, so the "legs" are even smaller (like a lightbulb filiment, but scaled down even more)
So, just like a lightbulb, the "legs" will eventually wear down from the voltage and break, severing the connections.

That is why overclocking hinders lifespan, because more voltage is being pushed through those already stressed out "legs" (and im sure if enough voltage was used, the transistor itself could burn out)

I see. I assume these "legs" are made of copper? Wonder if Intel could custom make one cpu where they are made from titanium! or gold! and it would never die. That would be pretty sweet.

The interconnects that connect the outputs from the silicon to the pins on the pin grid array, are made from pure gold, as gold has the best conductivity, these gold interconnects are hair thickness strands as you could imagine.

When gold?s temperature is brought down, down below freezing (-100 ish) it becomes a super conductor, which means it has even less resistance (Ohms), than it originally did, meaning that the flow of electrons can move alot faster without resistance, there are more variables when overclocking other than the yield of the die, for instance when you pump alot of voltage into these chips, there will be interference between the gold interconnects since i dont think they have any shielding against EMI, one of the reasons why they use low voltage (speed) and a higher ampere value, hence why the WATT consumption is so high on some chips (voltage x Amps = WATT).

which is why i think when people are using promie's and vapo's they are able to achieve far high overclocks, as they are able to reduce the resistance of the gold and further increase the voltage stably, as well as irradiating the heat out put from the silicon itself.

RichUK

 

[Mik3y]

personally, i dont care if overclocking shortens the lifespan of my cpu. omgwtfbbw, so what. instead of a long 10 yr life, it may only live to be 5 or 3 or 2. i'll have a new rig up and running by then.

 

[drifter106]

name something doesn't deteriate with age? Engines, light bulbs, human body. It's a process, same thing should hold true with the internal parts of a cpu. I have been told, that ever time a cpu is turned on and off a very,very, very small, almost insignifcantly part of the life of the cpu is lost.

 

[SamzAthlon]

I don't think it matters cus by the time the hoe is about to die, you'll have a new one in your hands.

 

[BigCoolJesus]

Originally posted by: RichUK

Originally posted by: Soviet

Originally posted by: BigCoolJesus
Yea, the transistors wear down over a period of time, and then break


think of it like this, a transistor on a cpu is ungodly small, so the "legs" are even smaller (like a lightbulb filiment, but scaled down even more)
So, just like a lightbulb, the "legs" will eventually wear down from the voltage and break, severing the connections.

That is why overclocking hinders lifespan, because more voltage is being pushed through those already stressed out "legs" (and im sure if enough voltage was used, the transistor itself could burn out)

I see. I assume these "legs" are made of copper? Wonder if Intel could custom make one cpu where they are made from titanium! or gold! and it would never die. That would be pretty sweet.

The interconnects that connect the outputs from the silicon to the pins on the pin grid array, are made from pure gold, as gold has the best conductivity, these gold interconnects are hair thickness strands as you could imagine.

When gold?s temperature is brought down, down below freezing (-100 ish) it becomes a super conductor, which means it has even less resistance (Ohms), than it originally did, meaning that the flow of electrons can move alot faster without resistance, there are more variables when overclocking other than the yield of the die, for instance when you pump alot of voltage into these chips, there will be interference between the gold interconnects since i dont think they have any shielding against EMI, one of the reasons why they use low voltage (speed) and a higher ampere value, hence why the WATT consumption is so high on some chips (voltage x Amps = WATT).

which is why i think when people are using promie's and vapo's they are able to achieve far high overclocks, as they are able to reduce the resistance of the gold and further increase the voltage stably, as well as irradiating the heat out put from the silicon itself.

RichUK

couldnt have put it better myself.......

The only thing that a Vapo pr Promie doesnt help is death my over voltage
ie, putting so much voltage through, that the CPU literally "burns out" (im not saying like 3V, but even 1.8 or 1.9 on a Venice is boarderline territory, whether you have aircooling or a Lquid Nitrogen, so there is a fine line)