• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Ex-Mining Card Maintenance

mjdupuis

mjdupuis

Member
I bought a 1070Ti which I know to have been used for mining, I don't know the extent (it couldn't be more than a year based on the release date of the card, Nov 2017) but let's assume moderate to heavy use for a year.

Should I remove the cooler and reapply thermal paste?
Would you overclock the card, or leave it at stock, assuming it's been heavily used?
Any recommendations for maintenance to ensure a long life?

Edited release date and spelling
 
Last edited:
mjdupuis said:
I bought a 1070Ti which I know to have been used for mining, I don't know the extent (it couldn't be more than a year based on the release date of the card, Nov 2018) but lets assume moderate to heavy use for a year.

Should I remove the cooler and reapply thermal paste?
Would you overclock the card, or leave it at stock, assuming it's been heavily used?
Any recommendations for maintenance to ensure a long life?
Click to expand...

Based on my experience as a miner, I'd check to see how the memory handles it. Most algorithms are memory intensive and tend to be heavily modified. The core, not so much in most cases. They are usually underclocked and undervolted. Run a game to see its stable without any artifacts.

If the card is suffering from degradation, no thermal paste changes will help it. Sometimes the degradation happens not just from overclocking, but from mishandling resulting in static discharge damage.

There's not much you can do other than how you'd normally treat a video card. Make sure it runs at a reasonable temperature, take precautions to minimize static damage, don't grab it by the PCB, rather by the sides or the backplate if it has one. Blow the dust out once in a while.
 
IntelUser2000 said:
Based on my experience as a miner, I'd check to see how the memory handles it. Most algorithms are memory intensive and tend to be heavily modified. The core, not so much in most cases. They are usually underclocked and undervolted. Run a game to see its stable without any artifacts.
Click to expand...

These are the details from the seller about how it was configured:
-100gpu, +550 mem for eth mostly. For zcash ~+100CPU, ~0 mem. Didn't mess with the fan too much, but never really got hot. Never over 75, probably around the 60s mostly.

+550 for memory seems sky-high
 
mjdupuis said:
These are the details from the seller about how it was configured:
-100gpu, +550 mem for eth mostly. For zcash ~+100CPU, ~0 mem. Didn't mess with the fan too much, but never really got hot. Never over 75, probably around the 60s mostly.

+550 for memory seems sky-high
Click to expand...
Those seem like normal, stable settings for mining - though he could have underclocked the core for more power savings. It's unlikely to have damaged the card. Just do some gaming stability tests, check that the fans aren't worn out, and clean out any dust. Unless it's overheating at stock settings, don't worry about the thermal paste. It's possible that the card might not OC quite as high as it would have when it was brand new, but it shouldn't be anything drastic.
 
Feld said:
Those seem like normal, stable settings for mining - though he could have underclocked the core for more power savings. It's unlikely to have damaged the card. Just do some gaming stability tests, check that the fans aren't worn out, and clean out any dust. Unless it's overheating at stock settings, don't worry about the thermal paste. It's possible that the card might not OC quite as high as it would have when it was brand new, but it shouldn't be anything drastic.
Click to expand...
Great, thanks for the perspective, I appreciate it.
 
IntelUser2000 said:
It may not have damaged the core, but you still want to test memory stability.
Click to expand...

I'm not aware of any specific way to test GPU memory, are there any?
I've got a bit of OC experience, I'm assuming gaming and stress tests will show any issues?
I don't plan to aggressively OC the card, but I'll take advantage of any headroom I have as this specific card seems to be overbuilt on the cooling (MSI 1070 Ti Titanium) with great temps even after OC.
 
mjdupuis said:
I'm not aware of any specific way to test GPU memory, are there any?
I've got a bit of OC experience, I'm assuming gaming and stress tests will show any issues?
Click to expand...

Yes, that's fine. Play some games with some memory overclock and see how it goes. I say real world applications are better testers than any benchmarking stress test. You could run a benchmark on top of that to make it better.
 
the biggest issue i've had with mining cards are fan failure.
the card temp may read a "nice level", that temp isn't the temp of the expelled heat which the fans get the full brunt off.

Most of the time i can just do maintenance on the fan, such as taking it apart and cleaning the metal shaft bearing, applying lube, assembling back together, and i'm up and running again.

On some fans, i replace the shaft bearing with a ball bearing.

on 5% of my cards, i have to replace the fans (which you can buy on amazon).

any fan problem i've had has always been non-oem fans, I've never had a blower fan fail or have issue with. However, there's a tradeoff. the blower fans sound like turbines. With most people, that is an issue, however for me, it's just white noise and helps me sleep at night. 😉

but for the time being, leave the thermal paste on, cause if you remove the heat sink, you might void any manufacturer's warranty that you might have.
 
This may or may not be relevant but I thought I'd throw it out there.
I took a 580 that had been used for protein folding 24x7 for 4 years.
I had a NIB 580 (both EVGA cards) and tested both each in the same machine running benchmarks and furmark/OCCT etc. for a weekend.
Both cards performed identically within margin of benchmark error and GPU temps were nearly the same.
Aside from the used one looking slightly "dirtier", there would be no way to tell them apart!
 
Just enjoy the card and play some games. As long as it's not packed full of dust, then the only thing you can do it replace the thermal paste. Hopefully you paid $250 or less for it so it's not a huge investment if something happens in a year or two. Also, which card is it? See if you can get it registered if something does go wrong. Evga for example will warranty it for 3 years from the manufacturing date if you don't have an original receipt.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top