cancel
Showing results for 
Search instead for 
Did you mean: 

Strange problem with my DUAL-RTX2080-8G

Nolan36
Level 7
Hi, first of all sorry if this isn't the right place to post (only for ROG products?),
but I'm extremely frustrated because I can't solve this issue with my DUAL-RTX2080-8G.
My Specs:
GPU: DUAL-RTX2080-8G
MB: AsRock Z390 Extreme4 Bios Version: 2.30
CPU: Intel Core i7-9700K
Memory: HyperX DIMM 32 GB DDR4-3200
PSU: be quiet! Dark P. Pro P11 850W
SSD: HyperX 960GB 480/550 Fury RGB (Windows & Software)
HDD: WD 2TB WD2003FZEX Black (only used for documents and large files)

In December 2018 I decided to finally replace my old PC because it wasn't upgradeable to win10. I configured and bought all the parts from the same retailer (Alternate; I'm located in Germany), assembled the PC myself and everything seemed to be running fine.
I was excited to finally play Battlefield 5 but the game would suddenly crash after random amounts of time (sometimes in menu, other times until 2 hours into a game session).
The crash itself would be a short freeze followed by a black screen and then recovering back to windows with the faulting application closed (in this case BF5). Once this crash had occurred the PC would be hit by black screens and recovering whenever I did something. So opening firefox or simply moving the file explorer window would result in this crash over and over again.
When I then proceed to restart the PC the problem would usually remain. Turning the PC off for 30+ minutes or taking it off the power and discharging it will solve the issue until another crash under high load occurs again.

Because this was a completely new build I couldn't pinpoint the issue to a single component; or if it was faulty software or hardware. After omitting Battlefield from my gaming the issue would start appearing during other situations as well; Monster Hunter World, League of Legends (although much rarer), Heroes of the Storm, Anno1800, Autodesk Maya and sometimes when being idle for a while and having a Youtube Video on pause. 3DMark only crashed twice with the error in at least 60+runs.

I have ruled out thermal being the problem because the card never runs over 80 degrees even when under 100% load. The problem seems to occur more often in 3D applications with higher settings. For example Anno 1800 on everything max won't last for 3 minutes, while turning the settings to medium or less would take 30 mins up to multiple hours without the problem.

Now the reason this took me so long to report is because I wanted to rule out as much as possible and I now believe only the GPU is left as the possible culprit.

What I have tried/done:
Windows reinstallation
Different Nvidia Drivers (removed with DDU), currently on the newest
Changing TDR delay (results in a hard crash without any recovering)
Different Driver/Windows Settings (Power plans are set to maximum & turned off all sleep options in windows as well as fast boot)
Different Monitor setups and inputs (Dual and Single monitor and HDMI, Display Port)
Switched PSU cables
Checked Memory and CPU
Underclocking the GPU
Different PC (my old win7 PC turned off completely after 5 minutes of BF1(also ruling out thermal/power))
Sending it to Asus via my retailer (took 5 weeks came back with a new bios but they found no problems)
Currently and during the last 2 months I've been using my old GPU without any issue or crash whatsoever

The Error shown in windows log when BF5 crashes:
Faulting application name: bfv.exe, version: 1.0.68.26495, time stamp: 0x5c092e76
Faulting module name: KERNELBASE.dll, version: 10.0.17763.134, time stamp: 0x1659a33b
Exception code: 0x887a0005

Error shown from Anno1800:
IDXGISwapChain::Present failed with error DXGI_ERROR_DEVICE_REMOVED.
GetDeviceRemovedReason returned 0x887a0006.

I have found no correlation between any stats from logs of GPU-Z during these crashes. My only observation is that it seems to occur more frequent when the clock speeds change around often (Boost Clock, Base Clock, Idle). As running 3DMark will hit boost clock and run under full load without problems during the whole test.

I have made a video of the constant crashing and recovering even consisting after a restart:
https://www.dropbox.com/s/20pv1zr8k7fsjb3/Sequence_Crash_RTX.mp4?dl=0

Does anyone have any idea what the cause of this could be and what else I could test? Otherwise my only option would be to sent the card in for RMA again but since they haven't found an error the first time I doubt the results would change.

Thanks for reading,
Nolan
9,141 Views
3 REPLIES 3

MrAgapiGC
Level 13
RMA the card. I know the issue.

Test as follow
1. game crash. The screen goes black and return or not or get stuck
2. on furthark is ok sometimes and then artefact detected.
3. drivers goes to windows dafult and stay there.
4. the screen is froze but the audio is there. faint but there. mouse are keyboard does not work
5. even restart does not restart until manually turn off PSU.

These are the issues that is comon on the asus card that i test in 2019.

Sometimes it manifest other a 1 week basis. crash and crach and again crash, but few days are ok.

IF these are you issues RMA THE CARD to the store and explain. I take the test on 6 2080/2080ti strix models.

ALl cards 2080ti/2080 are test before giving these to the client after purchase. And yes happens ON ALL CARDS FOR ALL BRANDS. RMA THE CARD. BE clear on what is what and what is not. since that is important so they can test the issue. one 1 card the errors manifest after 2 weeks after client explain what was the issue.


Cards that goes to 80C, on full load, is NOT cause of RMA. Those are drivers from nvidia ALONE since last 2 version.

Some errors WILL NOT manifest on test or 3dmark or firestrike and time to time are NOT artefacting., When they do, artifacts WILL appear on furtmark. since is the the only app that will not crash on test. IF that is present RMA the card. Be sure to put the fans on 100% before each test

NOTE: IS NOT THE MEMORY..... the the chip from the nvidia. Since is present also on samsung memory. these is why 2080super are here. So sorry I know there are more pro people on these. BUT IS NOT THE MEMORY!
Learn, Play Enjoy!

Didn't notice someone replied sorry.
Trying to get the Card RMA'ed but my retailer and the German Asus branch are a disaster.

To update on the problem:
Card is bad, seems to be something with the power unit on the card itself, it pulls Static power when the problem occurs.

More info on my Nvidia Support thread:
https://www.nvidia.com/en-us/geforce/forums/geforce-graphics-cards/5/325052/unknown-rtx-2080-error/

MrAgapiGC wrote:
RMA the card. I know the issue.

Test as follow
1. game crash. The screen goes black and return or not or get stuck
2. on furthark is ok sometimes and then artefact detected.
3. drivers goes to windows dafult and stay there.
4. the screen is froze but the audio is there. faint but there. mouse are keyboard does not work
5. even restart does not restart until manually turn off PSU.

These are the issues that is comon on the asus card that i test in 2019.

Sometimes it manifest other a 1 week basis. crash and crach and again crash, but few days are ok.

IF these are you issues RMA THE CARD to the store and explain. I take the test on 6 2080/2080ti strix models.

ALl cards 2080ti/2080 are test before giving these to the client after purchase. And yes happens ON ALL CARDS FOR ALL BRANDS. RMA THE CARD. BE clear on what is what and what is not. since that is important so they can test the issue. one 1 card the errors manifest after 2 weeks after client explain what was the issue.


Cards that goes to 80C, on full load, is NOT cause of RMA. Those are drivers from nvidia ALONE since last 2 version.

Some errors WILL NOT manifest on test or 3dmark or firestrike and time to time are NOT artefacting., When they do, artifacts WILL appear on furtmark. since is the the only app that will not crash on test. IF that is present RMA the card. Be sure to put the fans on 100% before each test

NOTE: IS NOT THE MEMORY..... the the chip from the nvidia. Since is present also on samsung memory. these is why 2080super are here. So sorry I know there are more pro people on these. BUT IS NOT THE MEMORY!


Didn't notice someone replied sorry.
Trying to get the Card RMA'ed but my retailer and the German Asus branch are a disaster and refuse to acknowledge a problem.

To update on the issue:
Card is bad, seems to be something with the power unit on the card itself, it pulls Static power when the crashing occurs,
as seen on the log:
https://drive.google.com/open?id=1-T3Q29Ic1-TQuViLGZkLWZC_NEFcPhHj

Had some people help me diagnose this on the nvidia support forums:
https://www.nvidia.com/en-us/geforce/forums/geforce-graphics-cards/5/325052/unknown-rtx-2080-error/