[SOLVED] GPU powers on, detected by BIOS, but not by OS

[auseklis74]

The GPU is ASUS ROG MATRIX 780 TI. Using it for several years.
How it started. Was playing game, suddenly some graphical artifacts showed up, screen freezes and PC restarted. So it happened in every game after few minutes of playing.
I turned on hardware monitoring on second screen and watched what is happening:
- GPU parameters are fine - temps are ok, clocks are ok;
- CPU cranks on all cores to 100 %, but temps are just fine - not more than 45 degrees Celsius.
After one day, my second screen showed no signal, main monitor showed picture. NVIDIA app did not started, and OS did not detected my GPU. My PC automatically switched to MoBo integrated GPU (I didn't knew it is possible, because monitors are plugged in GPU). No other app detects GPU (like afterburner etc)°
Checked BIOS settings - GPU is detected and recognized, and even temperature is read by MoBo. When PC is on, power connection LEDs are on, fans are spinning, side-panel logo is lid.
What I tried so far:
- plugged in other GPU in MoBo. It works fine, so there is no PCI slot issue;
- changed PSU to one that I don't have doubts that it works fine. Same result. So there is no power supplying issue;
- cleaned GPU from dust by using compressed air and brush (carefully) - no new results;
- triple checked that GPU is properly in slot;
- reinstalled NVIDIA drivers multiple times;
- reinstalled OS;
- in BIOS set priority to PCI graphics;
- turned on GPUs safe mode;
- tried this GPU on another PC - no results.
- tried to use monitor with HDMI cable (in this case monitor didn't even switched to MoBo integrated GPU).

Additional info:
- I have never tweaked GPU;
- Monitors are plugged with DVI cables (yes I know - old monitors!).
My PC specs:
MoBo - ASUS Z170 Pro Gaming
CPU - Intel i5 (6600k);
RAM - kingston DDR4 32 GB;
GPU - ASUS ROG MATRIX GTX 780TI
PSU - OCZ 750 W

I understand that GPU is a little bit deadish, repairing would take a lot of resources and I just have to let it go. But I would like to know if there is hope to wake it up, and definitely I would like to know what happened to my GPU?

 

[scout_03]

under control panel what you see for gpu driver does it use generic or not also you could use gpu-z rendering option to stress test gpu under load and check temp with others values on it with hwinfo .

 

[Satan-IR]

I agree with scout_03.

Usually when a graphics card starts acting up that is freezing or showing artifacts some time into gaming it's a matter of temps.

Those 45c can't be right for a GTX 780Ti under load. That card when under heavy load such as gaming should get around 70c.

I also recommend using GPU-Z and HWiNFO to see temps and info on card hardware. You can run GPU-Z, minimize it and run a game and when it starts acting up check the max temp with GPU-Z.

 

[auseklis74]

Problem solved!