CUDA Broke while running cuda app / cuda device not detected

I am running Windows 7, 16gb RAM, AMD Bulldozer cpu (8-core), Asus GTX 770, 3 monitors. I am running a cuda mining application, everything was working fine for the past 2 months. all of a sudden I logged in to the machine to do some work and my cuda app was still running but throwing cuda errors. closed the app down and restarted it and it says no cuda devices detected. well that’s strange… so I proceeded to download and install the latest GeForce drivers, no change, downloaded the latest cuda toolkit, no change. uninstalled cuda and all NVidia drivers and reinstalled all latest drivers, no change.

deviceQuery.exe Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 38
-> no CUDA-capable device is detected
Result = FAIL

I have been doing all of this via remote desktop connection and will have physical access to the machine tomorrow to diagnose it further.

I find it extremely strange that cuda would just break while the machine was running without being rebooted.

Any suggestions?

A WDDM GPU (any GeForce GPU) is not visibile to a RDP session. You’ll need to diagnose it directly on the machine or using some other remote connection such as Teamviewer or VNC. I can’t explain why it was working then quit, but maybe a reboot would have fixed it, but there’s no easy way to tell over RDP.

wow, thank you for the reply and you are complete and utterly correct. everything works fine when testing directly from the console.