[SOLVED] Problem with nvidia card or driver

[marekrz]

I have been trying to solve the graphics problem for some time. Once a week or more, an effect such this https://ibb.co/jZCcfqL https://ibb.co/FzCRWpz . The problem is not memory or processor - it has been tested.

There error appears in the event log with the source: nvlddmkm identifier 14

\ Device \ Video3

CMDre 00000002 000000c0 ff1fe203 00000007 00000000

The question is: what may be causing the problem and how to solve it? Please help me solve this problem. Thank you.

 

[Unolocogringo]

Your pic looks like memory errors.
Whether main memory or GPU memory is not evident.
How was the CPU and memory tested?
Was GPU memory tested?

Can you list FULL computer specs and temperatures while gaming.
Overheating memory could be the problem.

 

[marekrz]

Your pic looks like memory errors.
Whether main memory or GPU memory is not evident.
How was the CPU and memory tested?
Was GPU memory tested?

Can you list FULL computer specs and temperatures while gaming.
Overheating memory could be the problem.

Memory was tested by mdsched.exe
In fact I do not test GPU memory. I try to make this test. Thank you.

Problem not occurring during temperature or performance stress. In fact it never occurred during any game.

 

[Unolocogringo]

Test system memory with memtest86. let it run overnight if possible. I try for at least 5 to 7 passes of the full test suite
You can test video card memory with OCCT or FURMark.

Microsoft's memory checker is somewhat lacking at best.

You can also test memory with Prime 95 if you are sure your CPU is stable.
Run a torture test with blend settings. Errors will be memory related.

Might want to look at hardrive corruption as the next possible culprit. A failing hard drive when used as "Virtual Memory" can also cause this corruption pattern.

You did not list system specs so it could be as simple as cleaning the PCI-E slot and video card connector or memory modules and slots.
Components can creep over time with heating and cooling cycles and can be "FIXED" by removing and resetting them.

 

[marekrz]

Test system memory with memtest86. let it run overnight if possible. I try for at least 5 to 7 passes of the full test suite
You can test video card memory with OCCT or FURMark.

Microsoft's memory checker is somewhat lacking at best.

You can also test memory with Prime 95 if you are sure your CPU is stable.
Run a torture test with blend settings. Errors will be memory related.

Might want to look at hardrive corruption as the next possible culprit. A failing hard drive when used as "Virtual Memory" can also cause this corruption pattern.

You did not list system specs so it could be as simple as cleaning the PCI-E slot and video card connector or memory modules and slots.
Components can creep over time with heating and cooling cycles and can be "FIXED" by removing and resetting them.

Thank you very much. I try many hardware tests, but these tests found nothing. I have reinstalled Nvidia drivers and after it system worked fine for 3 weeks. But after Windows update last week and computer restart problem has back.