Telsa K80 CUDA driver crash Windows Server RC2 64bit

Hi

My GPU server keeps “losing” my GPU with the error “No CUDA devices found”.

From research I think this means the Nvidia driver has crashed?

This can occur after 10s of minutes or after a few minutes.

The only way I’ve found to fix it is to restart the server.

I’ve tried all versions of drivers from

http://www.nvidia.com/Download/index.aspx?lang=en-us

and

https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64&target_version=Server2012R2&target_type=exelocal

I’m connecting via Google Chrome Remote Desktop rather than RDP as I
know this causes issues.

I’ve also tried various fixes I’ve found online with no luck.

Do you have any ideas?

Thanks

overheating?

inadequate power delivery?

It was both! inadequate power because of a faulty fan. getting to 92 degrees!

Thanks for the hint :)