Screen crashing to random color

No i don’t have any other card . And yes the fan of the gpu spins like everything is working only the image on my screen is a stative color

Ps : it’s been more than a year that everything is normal but lately i’ve been getting this problem

Model number is 8300 covertible microtower

@saad-benhima

Can you try the GPU card in another PCI slot?

Just to prove that it may be hardware try running a standalone program and check if it happens with it running.

Running this program is not checking the RAM as such to see if it is faulty (even though it is doing that)it's just that it is a standalone program that doesn't need the OS to be loaded to be able to operate.

If the screen still fails when running it then it is definitely a hardware problem.

What is the model number of the desktop, you didn't say?

Have you checked that the latest BIOS version software and chipset drivers have been installed?

Maybe also check in Event Viewer for any Critical, Error or Warning events that are logged at the same time as the display malfunctions to hopefully get a clue as to what may be happening.

Being random and the GPU card also OK in another PC makes it difficult to trace as if it were there all the time it would be easier.

Also measure the PSU outputs by doing the paperclip test just in case they are lower or varying.

I tried the gpu in the other pci slot same problem occurs . For the bios , drives and the psu everything is fine .

I can’t do the test because when i turn on the computer it just freeze and the screen turns to a color .

Can the problem be from the pci slot or the motherboard ?

For the model number it’s an hp elite 8300 convertible microtower

@ saad-benhima

Can only suggest that you try starting the computer as a "barebones" PC and check what happens to try and nail the problem down further.

Disconnect the power cable from the HDD, SSD, ODD(optical disc drive), remove any other cards and only have the PSU, motherboard ( with memory), GPU card, keyboard, mouse and monitor connected and then start the PC and try to get into BIOS if the display is OK and check that it stays OK when in the BIOS menu screens.. Unless of course it switched to the Intel for BIOS menu displays

if you can do this try booting from the Memtest program and check as well

The reason for this is to check if the PSU is OK.

I know it seems that I'm concentrating on the PSU and that you said that the Intel GPU is OK, but this may mean that the Nvidia GPU is only operating at minimum power so less impact on the PSU. This is only to eliminate the PSU so that the problem is then on the motherboard. The PSU output may be failing under load so minimizing the load by disconnecting all "unnecessary" components (for testing purposes) may show something.

If it still fails as a barebones and you have done the"paperclip test" on the PSU then it will be the motherboard.

Finding out whether it is the PCI slots i.e. the PCI controller then as you said that it happened on more than one slot, or something else on the board would require the schematics for the board.

Maybe search online for "(insert motherboard 'board number') schematics" to hopefully find them. The board number is printed on the motherboard itself.

If you do find them then you will need to know how to read circuits and how to test with a DMM etc.