6 ms·

I certainly wouldn't want a second hand GPU that'd been used for mining; it'd be ragged.

viraptor 4y ago
Often repeated, but wherever you actually see someone testing that theory, they can't spot any issues.
deadbunny 4y ago
Is there any evidence that running a GPU/CPU at 100% for X years degrades the product vs average use over the same period?
- colejohnson66 4y ago
  No; there's been many tests. The mistake seems to come from thinking GPUs "wear out" like cars. The cooler might, but the silicon is fine. Some even suspect that crypto mining is better for the card than gaming as it avoids the constant ramp-up/ramp-down of the clock/activity by keeping it pegged at 100%. In turn, that would lower thermal stress.
  - remram 4y ago
    > GPUs aren’t cars I think another comparison is with hard drives, which are also used by some cryptocurrency schemes, and do degrade faster with intensive use.
    - colejohnson66 4y ago
      Yes, because hard drives, like cars, have moving parts that can wear out.[a] The silicon of a GPU doesn't have moving parts and is therefore more resilient. [a]: It's (hopefully) common knowledge in the tech community that purchasing used hard drives is a Bad Idea(TM)
      - the_gastropod 4y ago
        SSD's have no moving parts, and wear out too, right?
        w1nk 4y ago
        Yes, but for different reasons. The grandparent is probably incorrect though, there is emerging evidence that silicon is actually changing/failing over time. See this paper from google on their CPU cores where they have practical evidence of this occuring: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... If their data is correct, it should follow that these exact issues will happen on the small transistor process GPUs as well.
        throwawaylinux 4y ago
        It's not emerging, electromigration been a core constraint in semiconductor design for 60 odd years. E.g.,: Blech, I. A. and Sello, H. (1966). The Failure of Thin Alluminum Current-Carrying Strips on Oxidized Silicon. In Proc. Symposium on PoF in Electronics, 496-505.
  - w1nk 4y ago
    Actually there's reasonable evidence to believe this isn't true. This paper from google: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... outlines specific scenarios where CPUs fail over time. Given the evidence that these are silicon defects that are actually worsening over time, there's no reason to imagine these failures don't extend to GPUs as well. The difference in data here is obviously scale, google has -way- more CPUs than GPUs so the absolute counts of failures will be different.
    - colejohnson66 4y ago
      I'll concede that silicon can wear out over time; it's impossible for it to be immune from happening, and I wasn't speaking in absolutes. But as you mention, it's one of scale. I'm curious how likely a second hand GPU from a cryptominer is affected.
      - w1nk 4y ago
        So that's actually one of the awful implications of this paper. It's probably actually happening at a rate higher than would be noticed by humans. If a given piece of silicon is hosing up a GEMM (matrix multiply), in graphics scenarios this may be invisible to the human eye as it could potentially just introduce artifacts in a scene rendering that could be entirely ephemeral to the frame. In the case of crypto mining though, it's completely possible (probable?) that there are GPUs that can't possibly ever calculate a proper SHA3 hash (see the paper on AES instructions that fail in symmetric ways).
latchkey 4y ago
Totally untrue. I've run many many tens of thousands of GPUs for years, in the worst of conditions (directly exposed to seasonal temperatures and environment)... they keep on ticking. Power supplies on the other hand... those don't seem to last as long, but we open those up and they are hand soldered by someone in China.
- w1nk 4y ago
  It's boggling to me that everyone is so entirely convinced of this fact. Check out this paper from google where they observe this exact process occurring in their CPUs at scale: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... Given their conclusion that this is the result of manufacturing tiny process transistors, it follows that this can occur in GPUs as well.
  - latchkey 4y ago
    You're conflating workload related issues (which is what I responded to) and manufacturing defects / electron weirdness. Using GPUs for mining doesn't destroy them any more than using GPUs to render p0rn videos doesn't destroy them.
Melatonic 4y ago
Actually it can be the opposite - if the miner is using it properly and not overheating them regularly they can be in good shape. If they run at a high percentage but the temperature stays constant then theres not a lot of wear going on vs suddenly heating and cooling all the time. I imagine the bearings in the fan might have more wear, however, but that might be more easily fixed.

Often repeated, but wherever you actually see someone testing that theory, they can't spot any issues.

Is there any evidence that running a GPU/CPU at 100% for X years degrades the product vs average use over the same period?

colejohnson66 4y ago
No; there's been many tests. The mistake seems to come from thinking GPUs "wear out" like cars. The cooler might, but the silicon is fine. Some even suspect that crypto mining is better for the card than gaming as it avoids the constant ramp-up/ramp-down of the clock/activity by keeping it pegged at 100%. In turn, that would lower thermal stress.
- remram 4y ago
  > GPUs aren’t cars I think another comparison is with hard drives, which are also used by some cryptocurrency schemes, and do degrade faster with intensive use.
  - colejohnson66 4y ago
    Yes, because hard drives, like cars, have moving parts that can wear out.[a] The silicon of a GPU doesn't have moving parts and is therefore more resilient. [a]: It's (hopefully) common knowledge in the tech community that purchasing used hard drives is a Bad Idea(TM)
    - the_gastropod 4y ago
      SSD's have no moving parts, and wear out too, right?
      - w1nk 4y ago
        Yes, but for different reasons. The grandparent is probably incorrect though, there is emerging evidence that silicon is actually changing/failing over time. See this paper from google on their CPU cores where they have practical evidence of this occuring: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... If their data is correct, it should follow that these exact issues will happen on the small transistor process GPUs as well.
        throwawaylinux 4y ago
        It's not emerging, electromigration been a core constraint in semiconductor design for 60 odd years. E.g.,: Blech, I. A. and Sello, H. (1966). The Failure of Thin Alluminum Current-Carrying Strips on Oxidized Silicon. In Proc. Symposium on PoF in Electronics, 496-505.
- w1nk 4y ago
  Actually there's reasonable evidence to believe this isn't true. This paper from google: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... outlines specific scenarios where CPUs fail over time. Given the evidence that these are silicon defects that are actually worsening over time, there's no reason to imagine these failures don't extend to GPUs as well. The difference in data here is obviously scale, google has -way- more CPUs than GPUs so the absolute counts of failures will be different.
  - colejohnson66 4y ago
    I'll concede that silicon can wear out over time; it's impossible for it to be immune from happening, and I wasn't speaking in absolutes. But as you mention, it's one of scale. I'm curious how likely a second hand GPU from a cryptominer is affected.
    - w1nk 4y ago
      So that's actually one of the awful implications of this paper. It's probably actually happening at a rate higher than would be noticed by humans. If a given piece of silicon is hosing up a GEMM (matrix multiply), in graphics scenarios this may be invisible to the human eye as it could potentially just introduce artifacts in a scene rendering that could be entirely ephemeral to the frame. In the case of crypto mining though, it's completely possible (probable?) that there are GPUs that can't possibly ever calculate a proper SHA3 hash (see the paper on AES instructions that fail in symmetric ways).

latchkey 4y ago

Totally untrue. I've run many many tens of thousands of GPUs for years, in the worst of conditions (directly exposed to seasonal temperatures and environment)... they keep on ticking. Power supplies on the other hand... those don't seem to last as long, but we open those up and they are hand soldered by someone in China.

w1nk 4y ago
It's boggling to me that everyone is so entirely convinced of this fact. Check out this paper from google where they observe this exact process occurring in their CPUs at scale: https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s01-hochschild.pdf https://sigops.org/s/conferences/hotos/2021/papers/hotos21-s... Given their conclusion that this is the result of manufacturing tiny process transistors, it follows that this can occur in GPUs as well.
- latchkey 4y ago
  You're conflating workload related issues (which is what I responded to) and manufacturing defects / electron weirdness. Using GPUs for mining doesn't destroy them any more than using GPUs to render p0rn videos doesn't destroy them.

Melatonic 4y ago

Actually it can be the opposite - if the miner is using it properly and not overheating them regularly they can be in good shape. If they run at a high percentage but the temperature stays constant then theres not a lot of wear going on vs suddenly heating and cooling all the time. I imagine the bearings in the fan might have more wear, however, but that might be more easily fixed.