Black screen while gaming, fans 100% [logs included]

[nemfi]

Hello everyone!

Last 3 days i am trying to solve my problem which suddenly appeared after 2 years of Palit RTX 2080 Super Jetstream GPU ownership. While playing CS:GO or benchmark with 3dMark at different intervals a black screen appears and the fans of the video card spin to maximum, i still can hear audio and pc interface sounds, but only hard reset helps.

I tried many things as reinstalling drivers with DDU, replacing vga cables etc etc, nothing fixed my issue. Finally i tried my GPU on friend's pc with 1000W psu and faced the same problem.

What i already did:

• Tried DDU uninstall and install different GPU drivers
• Updated MB BIOS
• Changing GPU clock offset to -200 Mhz, Power target to 70%
• Maximum performance option in Nvidia control panel
• Switching to the GPU BIOS without factory OC (there physical switch between 2 kins of BIOS on the GPU) - Replacing 2x8pin VGA cables PSU -> GPU

Right now i logged sensors while crashing in 3dMark benchmark with HWiNFO64 and here it is (The exact time of crash is between 23:38:20 and 23:38:30, 816-820 rows in csv document):
https://docs.google.com/spreadsheets/d/e/2PACX-1vQsyYAsOKTWZFfoMAVn92lMZebfroa9G7L2huHay_ibGoGsGKbcG6sBJC7JWzSOTjq616imOty9EOrE/pub?gid=1362595003&single=true&output=csv
Also getting the warning in the event viewer: "Display driver nvlddmkm stopped responding and has successfully recovered"

I am grateful to everyone who could help in solving my problem

My configuration is:
OS
Windows 10 Pro 64-bit
CPU
Intel Core i7 8700K @ 3.70GHz Coffee Lake 14nm Technology
RAM
16,0ГБ Dual-Channel Unknown @ 1336MHz (18-18-18-43)
Motherboard
ASUSTeK COMPUTER INC. PRIME Z370-A (LGA1151)
Graphics
Optix MAG27CQ (2560x1440@144Hz)
DELL P2416D (2560x1440@59Hz)
PALIT RTX 2080 Super Jetstream (NVIDIA)
Storage
238GB Samsung SSD 850 PRO 256GB (SATA (SSD))
465GB Samsung SSD 970 EVO 500GB (Unknown (SSD))
PSU
be quiet! straight power 11, 750W


UPD:
I just noticed that, judging by the logs, right before the crash, GPU D3D Usage [%] and GPU Video Decode 0 Usage [%] increased from 11.7 to 95.5 and from 2.5 to 100.0 respectively, and then dropped to 0 and did not rise until the computer was restarted. Also GPU Memory Usage [%] dropped from 28.0 to 3.9.

UPD:

Also noticed the suspicious behavior of the indicators during the crash:

Read Activity [%]

Write Activity [%]

Total Activity [%]

Read Rate [MB/s]

Write Rate [MB/s]

0.3	0.6	0.9	0.038	0.265
0.6	0.6	1.3	0.003	0.056
19.8	1.0	20.8	20.359	6.310
5.9	0.2	6.1	7.416	0.388
0.0	0.3	0.3	0.000	0.193

UPD:

One more crash log (crash time about 2:47:56, row 440):
https://docs.google.com/spreadsheet...sITE/pub?gid=725612347&single=true&output=csv

 

[Mephisto510]

have you found a fix yet? I am having similar issues with my RTX2080ti. I have ordered a new power supply but I am also thinking of trying to replace the thermal paste on the GPU because I have found other posts where people have replaced their thermal paste and had success with this issue. I figure its a cheap easy fix and can't hurt to try. My system has been running fine for years. I put the 2080ti in last year and have never had issues. I used SPeccy and ran benchmarks and never saw the temp rise above 79c Also when running the benchmark the system didn't crash. But then later when scanning with MSI afterburner I had it crash again the same way with both monitors going black and fans running at 100%

 

[nemfi]

No, still didnt fix it. Also reball didn't solve an issue, PSU and thermal paste is ok in my case. Will try to change bios firmware idk...

 

[Mephisto510]

for me so far after running several GPU benchmarks and having the same issue happen every time I decided to replace the 16 pin PCI cable in my PSU for the video card. I didn't have another 16 pin cable but since I have a modular power supply I replaced the 16 pin with 2 8pin PCI power cables. I have ran several benchmarks and haven't had it crash yet so I have ordered a new PSU because I think that it may be starting to fail. I'm not sure exactly why the 2 8pins work. I am thinking maybe it doesn't cause such a power spike in the 1 module. Who knows maybe it was just the wire. I would rather get a replacement than to have it fail again and not have any other options. Especially since its going to take several days for the new power supply to get here.

the fact that you were able to reproduce the crash on your friends PC with the video card makes me think that it must be the video card. I would RMA it with Palit. I saw an article saying that they have extended the warranty to 3 years. Here is a link to the article. https://www.guru3d.com/news-story/gainward-palit-increase-warranty-three-years.html

 

[Hyderz]

yeah sounds like a hardware issue.. if its still under warranty get it replaced.
sounds like some solder somewhere is becoming loose?

before sending the gpu off could you try using your friends gpu to see its not other parts of ya pc.
just to eliminate the possibilities of the other parts of the computer.

 

[nemfi]

yeah sounds like a hardware issue.. if its still under warranty get it replaced.
sounds like some solder somewhere is becoming loose?

before sending the gpu off could you try using your friends gpu to see its not other parts of ya pc.
just to eliminate the possibilities of the other parts of the computer.

No way to replace it by warranty because seals are broken and BGA Reball made.
As i told, on friend's pc with my gpu same issue happens, so the problem is GPU, have to fix it somehow ):

 

[erocker]

No way to replace it by warranty because seals are broken and BGA Reball made.
As i told, on friend's pc with my gpu same issue happens, so the problem is GPU, have to fix it somehow ):

If the card has been reballed already, the GPU itself is probably defective. No fix for that.

 

[nemfi]

If the card has been reballed already, the GPU itself is probably defective. No fix for that.

Then why did it work fine almost 2 years?

 

[xman2007]

Then why did it work fine almost 2 years?

Because that's not a permanent/100% fix, it had issues before, and it has issues again now, it should have been repaired/replaced under warranty in the first place not reballed.

 

[nemfi]

Because that's not a permanent/100% fix, it had issues before, and it has issues again now, it should have been repaired/replaced under warranty in the first place not reballed.

Ofcourse, I would replace it if the replacement process did not take ~half of the year (computeruniverse you know). It was reballed few days ago. For 2 years it worked without problems.