I'm seeing high amount of "simulation became unstable" error on 970. Is this expected?

Can we expect an answer from scientists to whom we offer computing time? Thank you.

It's not expected. The question is if it's just one machine or systematic error.
From the daily WU report I see a 6% severe error rate for these WU's which is below our usual error rate, so I would say that it's machine dependent. With which I cannot help you much unfortunately. Maybe some other user in the forum can help you get the GPU more stable.

I'm seeing high amount of "simulation became unstable" error on 970. Is this expected?

Your GTX 970 is an overclocked one. I think you should raise the GPU voltage by 25mV to make your GPU a bit more stable.

I'm seeing high amount of "simulation became unstable" error on 970. Is this expected?

Your GTX 970 is an overclocked one. I think you should raise the GPU voltage by 25mV to make your GPU a bit more stable.

I wound up switching my nvidia-settings from "auto" to "adaptive" on the core frequency which seemed to eliminate the errors with the overclocked cards. I think some of the work units are overheating the overclocked cards. I take a little bit of throttle hit but nothing to write home about.