Question 6900 XT black screen crashes, screen tearing & stuttering ?

Toggle sidebar Toggle sidebar
R

rice0209

Distinguished
Oct 27, 2014
23
6
18,515
Hoping for some help from all the experts out there, thank you in advance!

I built a pretty nice setup here about 9 months ago. Here are the specs:

Asus x570-e mobo
Asus Tuf 6900 xt OC card (changed to liquid cooled - EKWB)
Ryzen 5900x CPU (Also liquid cooled)
32 gb G Skill Trident Z RGB (16gbx2) running at 3800mhz, FCLK 1900 (passed multiple memtests with 0 errors
2 TB Corsair M600 NVME
Corsair RM850x PSU (2 separate 8-pin cables connected to the 2 power inputs on GPU)
Windows 10, also was on Windows 11 with issues, before reverting to windows 10
Fully water cooled with fantastic water temps at all times - water is generally in the 23C to 26C range
GPU temps are generally around mid 40C with highest temp on card around mid 50C
CPU generally around mid 50s Celsius at worst

For 3 - 4 months or so, until around November 2021 or so, I never had any issues. Very stable. I mostly play Warzone but have experienced the crashes on other titles like Halo and COD Vanguard, but the bulk of my time is on Warzone when I play. Never had a crash outside of gaming.

What will happen, is at completely random times, my screen may start to flicker image, audio begin to cutout and then the screen goes black (lost signal). My computer will then try and reboot 2 times (i can tell by the noises, and led lighting transitioning), the monitor never regains signal. On the 3rd time, my computer will reboot back into windows. I was initially running some stable GPU overclocks for a long time. When the issues started happening, i have slowly dialed back any overclocks, including radeon settings in an attempt to isolate the issue. All i have done is been able to slow down the occurrences of black screen crashes, but have not yet eliminated them. As I posted above, I initially was running widows 10 when i first built, went to windows 11, started experiencing issues, did a couple complete wipes/installs back to windows 10 in attempts to clear any issues. Have tried every new AMD driver, including wiping with DDU, etc.

I have scoured the internet for this, finding similar problems and trying their solutions, but have not fixed it yet. Turned off practically all Radeon enhancements like sharpening, anti lag, enhanced sync. My GPU is running back at stock values, no OC. I will attempt to run my memory at 3600mhz, 1800 FLCK as my last OC setting left, but even at my current values (3800/1900) i get zero errors with memtest and multiple passes.

I am running out of ideas. I have seen posts recommending a higher wattage PSU but then also seem posts with a nearly exact build who tried a bigger psu but still had the crashes. I hooked up a watt meter to my PSU cord and never went about 600w at 120v, leading me to believe i haven't pushed any higher than that.

My event viewer has not been helpful at all with any critical errors other than the errors that seem to point to the failure in the past, not when it happened (like - The previous shutdown was unexpected).

If you have any suggestions for things i can look for to find what is causing this error, I would love to hear them. Hoping I do not have a faulty GPU. Would love t find a way to induce this problem and test it.

Thanks for your help!
 
Sort by date Sort by votes
Where did you source the installer for your OS? As for your motherboard, what BIOS version are you on at the time of writing? I'd advise trying out a different {SU with a higher wattage to rule out your unit being faulty or is failing. The thing with troubleshooting is that sometimes you need replacement parts to rule out your existing hardware is fine(and possibly the OS/drivers/software causing the hiccups).
 
Upvote 0 Downvote
Lutfij said:
Where did you source the installer for your OS? As for your motherboard, what BIOS version are you on at the time of writing? I'd advise trying out a different {SU with a higher wattage to rule out your unit being faulty or is failing. The thing with troubleshooting is that sometimes you need replacement parts to rule out your existing hardware is fine(and possibly the OS/drivers/software causing the hiccups).
Click to expand...

I bought my Windows 10 OS via Amazon in sealed Microsoft packaging on a usb thumb drive. I upgraded to Windows 11 via Microsoft online. When I had issues, I reverted back to Windows 10 (complete wipe/brand new install), again using my thumb drive.

I am going to overshoot and try a corsair HX1200 PSU. I already have the modular cables routed and tied up so this will be the easiest swap for the test. When my GPU was overclocked and using a lot of the Radeon enhancements I never had issues for months. But once it started stuttering and then going to black screen crashes almost nightly. Thats when I started backing down the overclocks and other settings. By the way, I also have turned off AMD's SAM (resizable bar) to keep things simple. I am now getting black screen crashes about once every week and usually only during my longest gaming sessions (say 3 - 5 hrs) but then every once in a while, i will get a crash after 1.5 hours. Just feels like no rhyme or reason to this.

As for my Bios, this problem has been happening on my x570-E motherboard over about 3 or so different bios updates over the last 6 months. I am currently on the latest and greatest bios from Asus, which is Version 4204 with Agesa V2 1.2.0.6b according to its notes.

I have heard that all AMD mobos will be getting an update through their manufactures to fix the fTPM issues sometime in May. I have almost given up with this black screen crashing problem, hoping this new bios may cure my problems since its a known major issue.

But as I said, I will try the new power supply and report back after enough testing time to feel more confident that it may be fixed. In the meantime, any further ideas are more than welcome.

Thanks!
 
Upvote 0 Downvote
Alright, so after installing the Corsair HX1200 Power Supply, after only 2 gaming sessions, i had the same problem. While playing Warzone, screen goes black/ monitor displayed no signal. I look at my motherboard and the q code is 9E which i believe is just normal operation / TPM is functioning. The computer begins an attempt at restart but with no graphics (white light comes on saying video error). This particular time, the computer appears to be running but still with no video. I use the power button to restart, and it restarts with the graphics card.

So i do not believe i have a power issue at this point.

I did get a windows event viewer error this time. Two actually, listed back to back. But it looks like they may have occurred during the first restart.

Event 18, WHEA-Logger

A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 0

The details view of this entry contains further information.


AND:


Event 18, WHEA-Logger

A fatal hardware error has occurred.

Reported by component: Processor Core
Error Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 20

The details view of this entry contains further information.

These were listed back to back, only difference being the APIC ID: 20 and 0


They occurred at 9:51 but I believe the initial crash then restart was at 9:50 according to the logs.

So confused as to what is going on here.
 
Upvote 0 Downvote
The issue is likely your Display Port cable, it needs to be 8K capable for adequate high refresh rate performance on higher end screens.

KableDirekt do a VESA certified 8K cable and they are not expensive.

I had this issue with my RTX 3070 and Radeon RX 5700-XT.
 
Upvote -3 Downvote
ProfSeaGoat said:
The issue is likely your Display Port cable, it needs to be 8K capable for adequate high refresh rate performance on higher end screens.

KableDirekt do a VESA certified 8K cable and they are not expensive.

I had this issue with my RTX 3070 and Radeon RX 5700-XT.
Click to expand...

I don’t think I posted my monitor in my build, my bad.

Asus VG27AQ - 1440 resolution at 165hz. Using the display port cable that came with the monitor.
 
Upvote 0 Downvote
rice0209 said:
Hoping for some help from all the experts out there, thank you in advance!

I built a pretty nice setup here about 9 months ago. Here are the specs:

Asus x570-e mobo
Asus Tuf 6900 xt OC card (changed to liquid cooled - EKWB)
Ryzen 5900x CPU (Also liquid cooled)
32 gb G Skill Trident Z RGB (16gbx2) running at 3800mhz, FCLK 1900 (passed multiple memtests with 0 errors
2 TB Corsair M600 NVME
Corsair RM850x PSU (2 separate 8-pin cables connected to the 2 power inputs on GPU)
Windows 10, also was on Windows 11 with issues, before reverting to windows 10
Fully water cooled with fantastic water temps at all times - water is generally in the 23C to 26C range
GPU temps are generally around mid 40C with highest temp on card around mid 50C
CPU generally around mid 50s Celsius at worst

For 3 - 4 months or so, until around November 2021 or so, I never had any issues. Very stable. I mostly play Warzone but have experienced the crashes on other titles like Halo and COD Vanguard, but the bulk of my time is on Warzone when I play. Never had a crash outside of gaming.

What will happen, is at completely random times, my screen may start to flicker image, audio begin to cutout and then the screen goes black (lost signal). My computer will then try and reboot 2 times (i can tell by the noises, and led lighting transitioning), the monitor never regains signal. On the 3rd time, my computer will reboot back into windows. I was initially running some stable GPU overclocks for a long time. When the issues started happening, i have slowly dialed back any overclocks, including radeon settings in an attempt to isolate the issue. All i have done is been able to slow down the occurrences of black screen crashes, but have not yet eliminated them. As I posted above, I initially was running widows 10 when i first built, went to windows 11, started experiencing issues, did a couple complete wipes/installs back to windows 10 in attempts to clear any issues. Have tried every new AMD driver, including wiping with DDU, etc.

I have scoured the internet for this, finding similar problems and trying their solutions, but have not fixed it yet. Turned off practically all Radeon enhancements like sharpening, anti lag, enhanced sync. My GPU is running back at stock values, no OC. I will attempt to run my memory at 3600mhz, 1800 FLCK as my last OC setting left, but even at my current values (3800/1900) i get zero errors with memtest and multiple passes.

I am running out of ideas. I have seen posts recommending a higher wattage PSU but then also seem posts with a nearly exact build who tried a bigger psu but still had the crashes. I hooked up a watt meter to my PSU cord and never went about 600w at 120v, leading me to believe i haven't pushed any higher than that.

My event viewer has not been helpful at all with any critical errors other than the errors that seem to point to the failure in the past, not when it happened (like - The previous shutdown was unexpected).

If you have any suggestions for things i can look for to find what is causing this error, I would love to hear them. Hoping I do not have a faulty GPU. Would love t find a way to induce this problem and test it.

Thanks for your help!
Click to expand...
Are you running with ray tracing on or off when it crashes? Have you tried running with default settings only and if so have you tried lowering the voltage? Maybe try the undervolt option in the tuning tab. I know that some people have had success when running ray tracing by lowering the voltage to 1.125 or 1.1. It may make running non RT titles more stable.
 
Upvote 0 Downvote
Third-Eye said:
Are you running with ray tracing on or off when it crashes? Have you tried running with default settings only and if so have you tried lowering the voltage? Maybe try the undervolt option in the tuning tab. I know that some people have had success when running ray tracing by lowering the voltage to 1.125 or 1.1. It may make running non RT titles more stable.
Click to expand...

No ray tracing.

it ran fine for a good 3 months with overclock on RAM, overclock on gpu (vram and core clock) with an undervolt.

then started having issues and have backed off any overclocking/undervolting on gpu first. Still had black screen crashes but at a lesser frequency so left gpu at default settings. Started tirning off any radeon enhancement options as well.

just recently backed of my RAM overclock as a last resort. Had it at 3800 mhz / 1900 FCLK. Always passed memtests at those settings. Dropped it down to 3600 mhz at 1800 FCLK. My memory is DDR 4000 and i am using the DOCP standard values for the ram.

hoping i can see if the memory change gives me any new insight.
 
Upvote 0 Downvote
I have not yet switched the DisplayPort cable because I was wanting to test my memory settings. I had previously ran at 3800mhz/1900 FCLK for a long while before the crashes/issues, and then even during the crashes as I tried many different changes to the system trying to find the culprit. My settings passed many memtests so i was thinking it was not the memory overclock since the GPU was crashing. But right before I got the new displayport cable in, I decided to reduce my memory speed down to 3600mhz/ 1800 FCLK. My gaming feels smoother and I have not seen a black screen crash after a handful of nights with a couple longer sessions. Beginning to believe that there was a very subtle instability that wasn't always hitting my computer. Maybe an issue that built up over longer sessions and wasn't detected by memtest?

So as of now, I am going with that as my issues as its the only variable I have changed this last go round that seems to be working. I wonder if at some point during one of the bios updates, it made my settings not quite as stable or introduced something that became an issue, as i said, I ran that overclock for months prior to my issues creeping up. I will at some point soon, try the displayport cable as well and see if I notice any difference in overall performance visually on the monitor.
 
Upvote 0 Downvote
Well, had another crash. So to do a quick recap, I did test a newer, larger power supply and still had crashes. I then backed down my memory overclock to 3600mhz for the ram and 1800 mhz for the FCLK. Had no crashes for a couple weeks and games were playing smoothly. Thought I figured it out and was having random instabilities with the higher overclock. So I then decided to switch out the display port cable, thinking that after I backed down the OC, i was good anyway.

A couple nights ago, had another crash. Same type of crash. Screen froze for about 1 - 2 seconds, then went black, audio was still trying to work but cutting in and out. Took about 10 - 15 seconds for the computer to finally fully crash and the computer attempted to reboot. Like usual, it failed its first reboot with no video. Then it rebooted but still had no video. Shut down and restarted with the front power button (not reset button) and the computer booted normally (with video) but with the same Radeon warning saying it defaulted to standard settings for the GPU because of the unexpected failure. As a note, I was not running overclocks on the gpu before this last week when I was trying to single things out. I only returned to OC'ing the GPU after I went a couple weeks with no crashes.

So after this crash a couple nights ago, I did check back with ASUS on their next bios update for the x570-e and they had a beta version of the new bios coming out in May which is supposed to address AMD's fTPM stuttering issues. I flashed the beta bios, then ran a memtest off a USB stick for over 4 hours overnight with 4 test passes. I had zero errors in that time with the 3600/1800 memory settings. Played for another 6 or so hours today with no crashes. But I am still not convinced it fixed it yet despite not having another crash since the bios update. So I thought I would see if anyone had any thoughts on what I can do. Still not getting any solid events in event viewer for either system or application errors. The crashed seem to surprise windows and leave no traces other than the restart information and that the crash was unexpected.

Is there anything I can run during a gaming session that might log more useful information and i can go back and check after a crash?

Thanks,

Chris
 
Last edited:
Upvote 0 Downvote
Any update on this one? I have exactly the same issue and tried countless potential solutions and still have that. It started suddenly and I could not get rid of...
 
Upvote 0 Downvote
If you want stable performance, you need to let your factory OC card run at default value's and settings, right after you have done a clean format and install.
If you keep playing with all the settings, you'll just run into trouble time and time again.

You may have already damaged your hardware by the sound of what you have been doing.
 
Upvote 0 Downvote
itrip said:
If you want stable performance, you need to let your factory OC card run at default value's and settings, right after you have done a clean format and install.
If you keep playing with all the settings, you'll just run into trouble time and time again.

You may have already damaged your hardware by the sound of what you have been doing.
Click to expand...

Factory OC card was reset to defaults a long time ago. The OC i ran was actually based off the numbers that ASUS uses in software when you use their OC. My card was watercooled and never gets remotely high temps. Not to say I couldn't have somehow damaged it at some point, but my OC's were very mild and always through the AMD software with its limitations. I wasn't bypassing any limits or using special software to do so.

All that aside, since the latest BIOS update from AMD via Asus, I have not had any crashes. My issue MIGHT have been from the fTPM stuttering issue this bios was aimed at fixing. I am back running my overclock that was stable for months and months prior to the crashing. I have also re-enabled SAM with no crashes. Keeping my fingers crossed that this bios update was the fix as I used to get random stuttering and eventually crashes once every two to three game sessions. Hoping I don't jinx this but I have not had a crash in probably 40 or more sessions. I did the beta bios update back in early or mid April then did the final release update that was released May 31. Haven't had a crash since the beta.
 
Upvote 0 Downvote
Apologies for not posting sooner but I have been tricked by this bug many times before and wanted to let this potential bios fix run for a while and see what happens. My issues definitely seem related to communication in the chipset between cpu, ram and gpu. Changing those memory settings helped alleviate and lessen occurrences/crashes. Wondering if months and months ago, when the fTPM was introduced if that's when it all started causing these problems. Reading across the internet, the effects of this fTPM issueon different system configurations have varied. Some with problems like mine and some with different issues. I wasn't going to try the workaround which was adding a new fTPM setup external from the board. Figured I would wait out the bios update.

My systems seems to be running a ton smoother now, was smoother than it was this past 5 or 6 months and is feeling more like it did originally when I first built it and had no crashes for the first 3 - 4 months or so.
 
Upvote 0 Downvote
I did get another crash (after the latest bios/fTPM fix). My crashes are few and far between now, even though it doesn't seem that the bios update fixed the issue. Since the only major change was removing the memory overclock, I am still leaning towards an issue with communication between ram, cpu and gpu. Sometimes, it has heavier stutters and sometimes it runs much smoother. When it last crashes, i wasn't experiencing a whole lot of issues of note, or any reason that would make me think a crash was coming on. I had not had a crash in so long (almost two months) that i was honestly shocked when it happened. At some point in the next month or so, I am going to take my system apart (water cooled) and check everything over. Thats the biggest downside to water cooling is the difficulty in pulling components to check for things like cracks in the motherboard or other oddities.
 
Upvote 0 Downvote
rice0209 said:
I did get another crash (after the latest bios/fTPM fix). My crashes are few and far between now, even though it doesn't seem that the bios update fixed the issue. Since the only major change was removing the memory overclock, I am still leaning towards an issue with communication between ram, cpu and gpu. Sometimes, it has heavier stutters and sometimes it runs much smoother. When it last crashes, i wasn't experiencing a whole lot of issues of note, or any reason that would make me think a crash was coming on. I had not had a crash in so long (almost two months) that i was honestly shocked when it happened. At some point in the next month or so, I am going to take my system apart (water cooled) and check everything over. Thats the biggest downside to water cooling is the difficulty in pulling components to check for things like cracks in the motherboard or other oddities.
Click to expand...
What bios release and driver release are you currently using? I've been reading up on more people having issues like you, but with all of the RX 6000 series cards. I'm starting to think it may all be related to the latest bios releases, because from what I can tell, people had more stable systems going back to last year before the September bios release.

If you're still getting crashes, you could try using a bios from before September if you haven't already.
 
Upvote 0 Downvote
+1 for count statistics with a lot of problems in RX 6900 XT and COD Warzone 🙁

My setup #1:
Gigabyte Z690 Aorus Elite AX
12900K
RX 6900 XT Sapphire Nitro+ SE

My setup #2:
Gigabyte X570 Aorus Master
Ryzen 9 5950X
RX 6900 XT Sapphire Nitro+ SE

I both setups, same problems. COD Warzone crash, freeze, reboots.

I noticed that the problem started a few months ago, because before it worked perfect.

Apparently it only happens in COD Warzone (the only game I play), but I use the GPU in Hackintosh but never had any problems.

I tested it on two monitors, with different DP cables (monitor's original, ugreen 8k and Vention 8k) - Samsung G9 49" 240Hz and Alienware 34" 120Hz.

I've formatted a few times, Windows 10, Windows 11 and several AMD drivers, all without success.
 
Upvote 0 Downvote
Third-Eye said:
What bios release and driver release are you currently using? I've been reading up on more people having issues like you, but with all of the RX 6000 series cards. I'm starting to think it may all be related to the latest bios releases, because from what I can tell, people had more stable systems going back to last year before the September bios release.

If you're still getting crashes, you could try using a bios from before September if you haven't already.
Click to expand...


Looks like I am running Asus bios 4403 dated May 31, 2022. I have been keeping up with their bios regularly since I had this problem never being more than a version or so behind. It very well could have been something introduced in the bios since I didn't have any issues until around the time I originally upgraded to Windows 11 last fall. I first thought it might have been the Win 11 upgrade, so i downgraded back to 10. Problem persisted. Then did two different clean wipes and reinstalls of Windows 10 trying to root out the problem but to no luck.

As for my drivers, I am on the 22.5.2 drivers from May 17th. Like the bios, I generally have kept up with the latest releases hoping for a fix.

After slowing my memory overclock down to 3600 with an FCLK of 1800, my crashes are way less frequent. But man, it will be sad if at the expense i put out for this motherboard, processor and memory, that it can't hold what most people with this same hardware would consider a mild overclock. My friends similarly built computer with the same ram but the Asus Tuf 570 board and 5800x processor runs this same overclock without issue. He has a RX6700 gpu. Its still weird that my computer ran so smooth, for months at 3800mhz/1900 fclk and then just magically started randomly crashing.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom