Display driver (AMD) suddenly crashing - any suggestions?

[dragonmike]

Been mining for over 2 months with this rig (7x RX570) without any significant issues.
1100/2000, 850mV core, 900mV mem controller.

Lately it had a great streak of about 300 hours without rebooting. Until yesterday.

Now, I can't even run Claymore for a minute without the display driver crashing. I've reduced memory overclock by 50MHz on all cards, no improvement whatsoever.

What would you say causes that sudden crashing?


EDIT: I am running Claymore 9.6, dual-mining ETH&DCR. GPUs are all Asus RX 570 Expedition OC with Elpida mem, running custom timings.

[1715126358]

1715126358

[1715126358]

1715126358

[1715126358]

1715126358

[Vann]

If you are using an external program to set the clock speeds, confirm what clock speeds the cards are running at when mining in Afterburner or HWinfo. I've had an issue where the clock speeds set in Claymore or Afterburner don't apply and the cards use the default clocks, which are too high for mining. Also try uninstalling the driver in safe mode using DDU and reinstalling.

[dragonmike]

You know what I think I just figured it out. It's always GPU 0, 2 and 6 that simultaneously go to zero, and then all locks up. These three GPUs (and the mobo) are on one and the same PSU.

PSU is dying, quite clearly.

It's the second EVGA 750 GQ that dies on me within 2 months of use. Go figure.

[Lipe490]

If nothing corrects it, try formating windows and put to windows 7.