What causes rig instability?

I made two rigs
Rig 1:
4 R9 280x
1 GTX 980
1 GTX 970
Rig 2:
1 R9 390
1 R9 290
1 R9 280x
1 R9 270x

Both rigs were stable until I purchased 2 new cards. I purchased 2 R9 390s and moved the cards around to this configuration to meet the power requirements.
Rig 1:
4 R9 280x’s - All on 1300 watt PSU
1 R9 270x
1 R9 290 - Both on 750 watt PSU

Rig 2:
2 R9 390
1 GTX 980 - all 3 on an 850 watt PSU
1 R9 390
1 R9 280x - Both on 750 watt PSU

GTX 970 was removed

After the cards were moved drivers were reinstalled because it was unstable. It is still unstable and was looking for some things that could have caused some instability. Rig 1 will run for a few hours then crash while rig 2 will mine for a few seconds then crash. I do not believe it is the risers because they worked fine for months before this unless they just now kicked the bucket. Any thoughts on how to go about troubleshooting or what could have caused this?

I’m going to try unplugging all cards and slowly add cards to until it crashes to maybe see if its a specific card that is causing the crashes. Do you guys think I’m drawing too many watts?

I would reformat the computers causing you trouble. Use win10 pro and shut off automatic updates. You may have a bad card as well.

I’ll go ahead and try that and report back here! Thanks for the recommendation. Both computers are causing me trouble so a fresh install of windows could hopefully fix it. Until then any more recommendations are welcome. I’ll just try them all once I get home haha.

It could be many things…

  1. You said risers worked fine for a few months… they may break in the mean time due to use or poor handling.
  2. Adding more cards means drawing more power, you either overload the power-supply or your circuit breaker. Temperature increase could also be a problem due to adding more cards.
  3. I even had an unstable rig due to a faulty power extension cord.
  4. Software, as @phakov said.

I will follow and reply if you have more problems.

  1. I can try swapping out risers but they did just work the day before I moved around cards. But like you said It could have been damaged due to poor handling.

  2. I did a rough estimate of how much power I would have been pulling and I do believe I have sufficient power. I again could be wrong. I’ll do some research to see how much each card pulls on full load (I dont have a Kill a watt… this could be a good investment haha).

  3. I’ll be sure to look over all my power connectors or extension cables. Good recommendation! This would be a thing that could have gotten overlooked.

  4. Yeah I’ll try formatting the drives and putting a fresh install of Win10 on there. I saw @phakov mentioned using the pro version and I’m currently using Win10 home. Would this maybe cause some instability?

Thank you both for the help. Ill be sure to try these ASAP.

You’ve never mentioned what the PSU wattage is for each rig, could you provide that information?

If you remove the additional GPU’s, does the rig return to stable?

Oh my bad! The 4 280x’s are ran by a 1300 watt PSU and the 270x and 290 are running on a 750 watt psu. On the second rig 2 390 and gtx 980 are running on a 850 watt psu and a 390 and 280x are running on a 750 watt psu. I can not answer that last question as my rig is in pieces right now trying to troubleshoot haha. I’m starting to thing it was just a riser. I may have had multiple dead risers I was testing it with. I’ll report back here when I figure out the riser situation…

So! I tested it and the one card Is confirmed dead. When I put it as my main GPU there are lots of graphical errors and then it crashes. RIP R9 390 :worried:

Edit: The person I purchased the 390 from told me to come back to return it and he will replace it with a R9 390x so free upgrade. Just glad this guy was nice about it!

1 Like

Wich mining software do you use on these “mixed” rigs??

DSTM for NVidia cards and Claymore for AMD cards.

For the second rig I would try to leave out the gtx prhaps it causes problems to run two miners over the same pcie lanes? -I’m no expert…
And for the first one you could try to undervolt and first let run by stock settings for gpu- & core clock so you are sure that you didn’t overoverclocked something.

Once I removed the dead card that seems to have fixed the problem. If the instability continues Ill keep these in mind. Thanks!

Sounds good…:+1:(20characters)

Glad you found it, I guess i am about to do the same. I think its a bad 1070ti causing my crashes. :frowning: My nicest card in my rig. Hope its not it.

Yeah I really hope its not a dead card either for your sake. I was able to get my card replaced for free so that was nice.

Both Claymore and EWBF allow the user to turn off (not mine) on certain GPUs. How to do this is that readme file. When I suspect a bac card, I shut it off with the software and wait. If the crashes stop then I pull the card and test it alone to pin it down. If it’s bad into the bin it goes :frowning:

1 Like

Or you can do what any sane person would do like linus tech tips… Put it in a toaster oven and try to fix it! :joy:
But in reality, your way of troubleshooting would be a good way to troubleshoot cards. For me at least it seems like the most likely cause of a card acting funky is a dead riser. I’m not sure if its that I just didn’t get good risers or what but I had… I think 2 or 3 die on me. They made things like hashrate dips and the miner just all together crashing occur. What risers do you guys get? I should probably start buying better quality ones.