This GPU: https://www.techpowerup.com/gpu-specs/g ... -780.c1701

Hello,

I had Artifacts and vertical lines on my GTX 780 graphics card. When I tested it with Nvidia Mats, the A1 ram appeared to be faulty on my GTX 780 card. I replaced it with a completely new ram. There are no artifacts and the card is fixed, but PC shutdown with the Kernel-Power 41 error. I enter the game and play for about 1-2 hours. At random times the computer freezes with a black screen and windows sound. I have to shut down the computer. I tested with the programs below, but I did not experience any shutdowns!

Furmark -- 35 Minutes
MSI Kombustor 4 x64 -- 1 Hour
Valley Benchmark 1.0 -- 15 Minutes
Superposition Benchmark -- 25 Minutes

When I do it with Nvidia MATS/MODS, I get Error 33. However, when I start the test with -no_scan_out I do not get any errors! Also, when I start the test again after the test, I do not get any errors.

What are your thoughts on this subject?

I have no knowledge of Russian language. I'm sorry for that.

Hi. I apologize in advance, I'm not very good at English and I translate through a translator.

The symptoms are similar to overheating of the "VRM zone", check how the mosfets are cooled. I had a similar situation on the GTX 770 from ASUS, but I had a couple of mosfets broken. That is, the card worked, passed all the tests, but with some games the screen turned off and the sound remained in the background. At first, I solved the problem by replacing thermal pads, but it didn't last long. After reducing the power of the video card through MSI Afterburner, the shutdowns disappeared. As a result, I replaced the faulty mosfets and everything began to work well

I also met a similar problem on the RX580 and 1080, but everything was fine there, the VRM was just overheating very much.

As you said, I made the following settings with MSI Afterburner and it did not shut down for 5 hours today!

https://imgur.com/RPwD5Az

I will check the mosfets. Thank you very much for your reply !


I noticed that there is not one SMD in the RAM. Could it be this?

https://imgur.com/70NVvFI

I'm glad it helped you. If your settings stop working, then look at how I see it - I reduced this parameter, set it to 80 percent, and it helped me https://imgur.com/a/eUEUQaQ

No, the absence of this capacitor should not cause such consequences. But for the sake of experiment, try to install a capacitor, you have a similar circuit at the bottom

I installed capacitor but the problem still persists. Msi settings don't seem to work anymore. When I set the game to high settings, the computer freezes with remix sound after 20-30 minutes, sometimes 1-2 hour.

I tested the mosfets and they are fine.

Stock vram 310. I installed (401) vram.What exactly did 310 mean?

https://imgur.com/a/d4wlx6X

Also, for testing purposes, a standart 470uf 16 capacitor was installed instead of 270uf 16v. Could this be the reason?

https://imgur.com/JivG7cg

The screens randomly just turn solid color. Random solid colour(White,Yellow,Grenn) screen freezes.

Probably, this was bad idea. Faulty 7** gpu's behavior is still affected by heat. It may work normal for some time, then issue comes back.

I know 2 ways out:
1. Warm GPU out. This method was widely known, and there is many ways to do that. I'm using 100-160C at 15min. But effect doesn't lasts long any way.
2. Just disable faulty chanel by tool from THERE. This method was published not so long ago, and it should give way more stability.