Claymore CUDA error with 7th card at startup but not all the time

[KevinUK]

Hi,

I have 7 x GTX 1070 and MSI Z170A motherboard not overclocked and a virtual memory of 16gb.

There are no problems running 6 cards, I have ran that all week.

When I run 7 cards, I have had it running for over 48hrs before I needed to reboot for unrelated things.

If it starts mining on 7 cards then it runs fine.

The problem is when I start the miner, I nearly always get stuck in a loop of DAG errors.

If I leave it running in this state for 5 minutes then it sometimes fixes itself.

Do you have any solutions to not get into this state to begin with? I've spent a few hours unable to get 7 cards running again.

https://i.imgur.com/DTdUiDo.png

https://i.imgur.com/sEDSNvG.png

Thanks!

[family2806]

For multi-GPU systems, set Virtual Memory size in Windows at least 16 GB:
"Computer Properties / Advanced System Settings / Performance / Advanced / Virtual Memory".

[KevinUK]

Forgot to mention that I have 16GB set already.