Do cpu coolers go bad?

[testcy]

Do cpu air coolers ever go bad? I am asking about the heatsink, not the fan. The fan can become noisy or stop working completely, what about the heatsink?

[filpo]

Just now, testcy said:

Do cpu air coolers ever go bad? I am asking about the heatsink, not the fan. The fan can become noisy or stop working completely, what about the heatsink?

No they generally won't because they're essentially big hunks of metal, they can perform worse if they get really dusty or not much airflow but no the heatsink doesn't go bad. Thats why you see decent cooling on some 15 year old machines that are still respectable

[Bismut]

Theoretically, yes. As @TechlessBrosaid, heatpipes can fail, but I have never heard of that happening and an air cooler with two fans is the gold standard for reliability (other than dust buildup, for which a passive cooler would be best - highly unlikely in a Server or so though)

[Bombastinator]

1 hour ago, testcy said:

Do cpu air coolers ever go bad? I am asking about the heatsink, not the fan. The fan can become noisy or stop working completely, what about the heatsink?

Yes but it takes decades.  Everything wears.  Eventually the moving water in the tubes will wear away the scintered metal beads the wall of the tube is lined with.   The power of flowing water is quite wild. It’s just not a fast thing.

[leclod]

41 minutes ago, Bombastinator said:

Eventually the moving water in the tubes will wear away the scintered metal beads the wall of the tube is lined with.   The power of flowing water is quite wild. It’s just not a fast thing.

I wonder how much wear is, in those cases, due to minerals and other stuff caried in the water. I guess a lot.

And I guess there's no such things in that distilled pipe water. So wear would be minimal other minimal.

[Somerandomtechyboi]

A chunk of solid metal degrading? Nah, maybe the fan(s) will die every decade and dust buildup but thsts about it for maintenence

 

An aio? Those will die in 5-7 years usually pump failiure but the water inside can also just evaporate through the tubes

[Bombastinator]

3 hours ago, leclod said:

I wonder how much wear is, in those cases, due to minerals and other stuff caried in the water. I guess a lot.

And I guess there's no such things in that distilled pipe water. So wear would be minimal other minimal.

Yep.  Very minimal. Non zero though.  I wouldn’t be at all surprised if an air cooler does a century or two.  Nothing lasts forever.  Some things can last a long time though, and air coolers are one of them.

Edited March 26, 2023 by Bombastinator

[Lurking]

On 3/26/2023 at 3:27 AM, TechlessBro said:

Heat pipes can fail in rare cases.

But not the solid copper or alloy fin stacks.

They have up to 6 heatpipes. Even if one heatpipe leaks out the refrigerant, you wouldn't notice.

 

Even if a fan fails, a large air cooler has enough passive cooling capacity to run Windows. Just no heavy duty loads.

 

Water coolers on the other hand....

[freeagent]

If you lost a pipe you would notice for sure. The only way you would lose a pipe is through a defect, not impossible, but highly unlikely.

[Lurking]

8 hours ago, freeagent said:

If you lost a pipe you would notice for sure. The only way you would lose a pipe is through a defect, not impossible, but highly unlikely.

The ends are soldered or crimped and that could leak.... like 1 in 1000,000,000.

Or the interior coarse wicking material could react with the refrigerant (unlikely since I assume they all use similar material and refrigerant).

 

But one of 5 pipes failing wouldn't necessarily be noticeable unless we talk about systems working at the edge of the cooling design limits. 

[freeagent]

5 minutes ago, Lurking said:

The ends are soldered or crimped and that could leak.... like 1 in 1000,000,000.

Or the interior coarse wicking material could react with the refrigerant (unlikely since I assume they all use similar material and refrigerant).

 

But one of 5 pipes failing wouldn't necessarily be noticeable unless we talk about systems working at the edge of the cooling design limits. 

There are photos of popped heatpipes on google somewhere. If it’s a 6mm pipe it’s about 40w you lost, if it’s an 8mm pipe it’s about 60w..

[Lurking]

16 minutes ago, freeagent said:

There are photos of popped heatpipes on google somewhere. If it’s a 6mm pipe it’s about 40w you lost, if it’s an 8mm pipe it’s about 60w..

Each added pipe has diminishing returns. If you have 5 pipes, one pipe may provide 30% of cooling. Adding the second may add 20% and so on and going from 4 to 5 may only give you 10% more. So you only lose so much. 

 

I'm sure there are photos, but it isn't a widespread issue like water AIO failures. 

[TPCEA]

As was pointed out, the water may damage the interior of the pipes, which would take a very long time and is highly unlikely unless the manufacturer was eminently stupid and/or greedy. Pipes can break open, too, but that's also rare. Or, the water can be vaporized to the point of the tube being dry, which will make it overheat. I think that unsoldered heatsinks are more vulnerable to damage, and I imagine that solder can fail, especially when applied poorly. Aside from that, user error like not removing the plastic and not using paste might contribute. Still, all of these things are very rare unless you buy a poor-quality cooler. 

[thrasher_565]

2 hours ago, Lurking said:

Each added pipe has diminishing returns. If you have 5 pipes, one pipe may provide 30% of cooling. Adding the second may add 20% and so on and going from 4 to 5 may only give you 10% more. So you only lose so much. 

 

I'm sure there are photos, but it isn't a widespread issue like water AIO failures. 

its because the heat source in moving away for the heat pipe. the his dose help move the heat but its not perfect. and really depends on cpu. as far as i under stand it they start out with the fastest cpu and go down from there by down clocking  them or disabling cores and thows core can be random. hence the silicon lottery. because the lower the nm the more fail rate. 

 

but the stuff in the tubes can dry out, leak and then with cpus getting bigger might not fit on the cpu nor no mounting option. of cores there diy thow.

 

not only that they find ways to make cou cooler better thow at a slow rate... the newest cpu thoe need all the cooling kinda like 8700k all over again

 

the heat sensors in the cpu are also unknown witch can make temp vary.

[Lurking]

5 minutes ago, thrasher_565 said:

its because the heat source in moving away for the heat pipe. the his dose help move the heat but its not perfect. and really depends on cpu. as far as i under stand it they start out with the fastest cpu and go down from there by down clocking  them or disabling cores and thows core can be random. hence the silicon lottery. because the lower the nm the more fail rate. 

 

but the stuff in the tubes can dry out, leak and then with cpus getting bigger might not fit on the cpu nor no mounting option. of cores there diy thow.

 

not only that they find ways to make cou cooler better thow at a slow rate... the newest cpu thoe need all the cooling kinda like 8700k all over again

 

the heat sensors in the cpu are also unknown witch can make temp vary.

That's why you get a cooler with a solid block instead of exposed heatpipes. That block tranfers heat between pipes. This also works better with the modern CPU that have chiplets on one side instead of the center.

 

With exposed heatpipes you rely a lot on the heatpipe directly over the chiplet. A block basically is a thick heatspreader.

[thrasher_565]

11 minutes ago, Lurking said:

That's why you get a cooler with a solid block instead of exposed heatpipes. That block tranfers heat between pipes. This also works better with the modern CPU that have chiplets on one side instead of the center.

 

With exposed heatpipes you rely a lot on the heatpipe directly over the chiplet. A block basically is a thick heatspreader.

ya alot of cpus work for the cheap cooler but now we needing more cooling probably see that on more expensive coolers. it was a way to cut cost but that most likely means even higher cost for coolers...

 

we all no what happens when the heap pipes are not flat like on the corsair cooler...

 

i have an old thermltright 120 ultra cooler and it has the blocked heat pipe and funny enough a mount was made for it so i could use it on other sockets instead of 775 and 1366 witch would be cool to test an old top of the line cooler...

 

but to be fair it was a $100 cooler at the time..

Edited March 28, 2023 by thrasher_565

[Bombastinator]

I suspect that it will be a situation of “really?! That’s a computer?!  Like from the olden days or something?!” Long before one of them wears out.  Pretty much every other component will wear out before an air cooler does.

[Beerzerker]

1 hour ago, Bombastinator said:

I suspect that it will be a situation of “really?! That’s a computer?!  Like from the olden days or something?!” Long before one of them wears out.  Pretty much every other component will wear out before an air cooler does.

Can attest to that.
I have coolers that's over 20 years old here that, to this day still work fine and are perfectly useable.
Occasional damage to the fins for example is one factor but most of the time that can be fixed easily. As long as one is careful with their cooler it will outlast the machine it's in with ease, only issue that may happen are the fan(s) going bad over time and that's natural to happen.

Do note older ones with heatpipes have a block the heatpipes are bonded/attached to on top or it's sides, newer style coolers really don't have a base so much with heatpipes being part of the cooler base and CPU contact surface of the base itself.

So no - In actuality they don't "Wear out".

[Bombastinator]

3 hours ago, Beerzerker said:

Can attest to that.
I have coolers that's over 20 years old here that, to this day still work fine and are perfectly useable.
Occasional damage to the fins for example is one factor but most of the time that can be fixed easily. As long as one is careful with their cooler it will outlast the machine it's in with ease, only issue that may happen are the fan(s) going bad over time and that's natural to happen.

Do note older ones with heatpipes have a block the heatpipes are bonded/attached to on top or it's sides, newer style coolers really don't have a base so much with heatpipes being part of the cooler base and CPU contact surface of the base itself.

So no - In actuality they don't "Wear out".

This may be a question of definition of terms.   EVERYTHING wears out.  Eventually.  It may be a question of geologic scales  of time, but it will happen.  Iirc the guy who made Mount Rushmore put an extra inch on the top of Washington’s nose, because rain.how long does it take rain to wear an inch of granite?  A long time I would guess.