Gray screen crash when gaming, what could be the issue?

[emothxughts]

Specs:
CPU: AMD Ryzen 3 3200G
CPU cooler: Deepcool GAMMAXX 400 V2 64.5 CFM CPU Cooler
Motherboard: Asus PRIME B450M-K II, BIOS American Megatrends Inc. 3802 (04/28/2022)
Memory: PNY Electronics 8192 MB (DDR4-2133) - XMP 2.0 - P/N: 8GBF1X08QFHH38-135-K x2
Storage: PNY CS900 960GB SSD (FW: CS900613), HP SSD EX900 500GB (FW: T1102A0L) (the boot drive)
Video Card: Gigabyte Radeon™ RX 5600 XT WINDFORCE OC 6G (2 fans version), driver 31.0.12029.7019 (Adrenalin 22.11.1) DCH / Win10 64
Case: Gaming Freak S920 PRISM
Power Supply: Cooler Master MWE Bronze V2 650W (230V non fullrange), PSU tierlist C-tier
Monitor: LG 24MP400
More details at https://valid.x86.fr/75748p

I've been having gray screen crash problems with my setup lately. Randomly, while gaming, the audio will stutter and stop and then is followed by the screen turning gray, like when you start a game, not a black screen like when display output is off.

Indeed, when I pull the HDMI cable from my card, the display turns black followed by a no input message as expected, then when I plug it back in, the display turns back on, with the gray screen, as if the card is still putting out a signal, though that signal is pure black-gray nothingness.

The system is completely locked up and non-functioning, though my case fans, CPU cooler fan and graphics card fans are still spinning, with lights on, and NOT at 100% speed. Only thing I can do then is force shut my PC by long pressing the power button.

I stress tested my CPU & RAM with OCCT and Testmem5 overnight with no errors, ran Heaven & Superposition with no crashes, and I can browse the web and run non-game tasks with absolutely no problems, but when I play games, it's only a matter of time before a gray screen crash occurs, in games like Elden Ring, Skyrim SE and recently Sleeping Dogs. And in the case of Elden Ring, it gray screened right when it starts loading.

I monitored CPU & GPU temps with OCCT and Radeon Software, they're not overheating either. (Typically CPU is 51 degrees C and GPU 70+ degrees C). I also checked the Event Viewer for errors, the only errors are Kernel Power 41 (63) errors, which I'm pretty sure is just the result of me long pressing my power button to turn my PC off. I also reinstalled graphics drivers, DDU-ing every time beforehand, to no avail.

That said, I also remember having XMP turned on when these crashes occur, as well as having OBS installed and running.

So please, I would appreciate it if anyone can figure out what's wrong with my PC, why it crashes to a gray screen when I'm gaming. Is it the GPU, the RAM, the PSU, or something else entirely?

 

[neko77025]

Not sure if this is it or not .. but give it A try. Had me stumped for MONTHS.

(can read about my issues here , everything I tried including MANY RMAs...)

(What Would you do) Computer Constantly Crashes ... GPU ? CPU ?

Ok .. So I am 100% confused at what is causing my issue. .. It down to Motherboard , CPU, GPU or Maybe PSU? ( I know , i know that almost everything) So about 3-4 months ago My PC started to Randomly reset or Blue screen ( whea_uncorrectable_error ). Thing is , it was so rare it was hard...

www.techpowerup.com

But long story short .. was Xbox Game Bar ... Turn it off V

 

[emothxughts]

Not sure if this is it or not .. but give it A try. Had me stumped for MONTHS.

(can read about my issues here , everything I tried including MANY RMAs...)

(What Would you do) Computer Constantly Crashes ... GPU ? CPU ?

Ok .. So I am 100% confused at what is causing my issue. .. It down to Motherboard , CPU, GPU or Maybe PSU? ( I know , i know that almost everything) So about 3-4 months ago My PC started to Randomly reset or Blue screen ( whea_uncorrectable_error ). Thing is , it was so rare it was hard...

www.techpowerup.com

But long story short .. was Xbox Game Bar ... Turn it off V

Thanks, I'll try disabling the Xbox Game Bar. That said, since the original post, I disabled my XMP/DOCP too.

Some more details:

• I did reset my Windows 10 2 weeks back. Before that, I get SYSTEM_SERVICE_EXCEPTION and PAGE_FAULT_IN_NONPAGED_AREA errors trying to play The Witcher 3. In fact, that was what made me reset my Win 10 in the first place. Now after the reset, I don't get BSODs anymore, but then I get the gray screen errors described in my original post.
• I tried the Windows key+ctrl+shift+B combo when it grayscreens, no beeps or changes in the display like it should in normal operation. No dump files either for WhoCrashed to read.
• Save games made in Skyrim SE when the crash occurs are corrupted, cannot be loaded, but I can load saves before the gray screen just fine.
• My speakers (cheap dual speakers, 3.5mm jack type) are static-y when in normal operation, I can hear beeps when I move my mouse around when in normal operation too. (Not a big problem, got used to it.) When it gray screens though, the static is gone, and so are the beeping noises from moving my mouse.

 

[neko77025]

Thanks, I'll try disabling the Xbox Game Bar. That said, since the original post, I disabled my XMP/DOCP too.

Some more details:

• I did reset my Windows 10 2 weeks back. Before that, I get SYSTEM_SERVICE_EXCEPTION and PAGE_FAULT_IN_NONPAGED_AREA errors trying to play The Witcher 3. In fact, that was what made me reset my Win 10 in the first place. Now after the reset, I don't get BSODs anymore, but then I get the gray screen errors described in my original post.
• I tried the Windows key+ctrl+shift+B combo when it grayscreens, no beeps or changes in the display like it should in normal operation. No dump files either for WhoCrashed to read.
• Save games made in Skyrim SE when the crash occurs are corrupted, cannot be loaded, but I can load saves before the gray screen just fine.
• My speakers (cheap dual speakers, 3.5mm jack type) are static-y when in normal operation, I can hear beeps when I move my mouse around when in normal operation too. (Not a big problem, got used to it.) When it gray screens though, the static is gone, and so are the beeping noises from moving my mouse.

yea , i was having no dump files with mine .. If it still happen not sure not sure after that .. if it stops after you try both .. turn one back on to see if it comes back so you will know what did it.

 

[emothxughts]

I'll try playing a game again. Still anxious about gaming after the last grayscreen.

That said, what does it mean when I'm still able to get a signal out after I plugged in my HDMI cable to my card again (albeit with just the blank black-gray screen I'm describing the whole time)? Doesn't say "no display output" on my display unless I actually unplug my cable.

 

[emothxughts]

Update:

I previously had a Ryzen 5 3600, newly bought, that I had to RMA because it kept black screening and restarting my PC, throwing WHEA 18 Cache Hierarchy Errors, and I was running that RX 5600 XT with it. I received my replacement R5 3600 yesterday, installed it with all BIOS settings at default (so no XMP/DOCP), stress tested with OCCT (Large) with no problem. Then when I played a game today, a fairly GPU intensive one, my PC blackscreened, audio glitched like in the original post, reboots and I get the same WHEA 18 error with the previous R5 3600.

I recall running my spare Sapphire Pulse RX 550 card with the R5 3600 and having no problems whatsoever, besides it being a weak GPU of course. I'm now beginning to think that the real problem, the weakest link, is the RX 5600 XT, and that the grayscreen problem with the R3 3200G and the reboot with WHEA 18 with the R5 3600 are both caused by the same borked thing.

I returned the RX 5600 XT today for warranty, and I'm currently running my older Sapphire Pulse RX 580 4GB. The old workhorse ran relatively fine with my R3 3200G with XMP on, though it does blackscreen occasionally, not as frequently as the possibly borked RX 5600 XT though.


TL;DR: Installed RMA replacement R5 3600 with all BIOS settings set to default, I restart with a WHEA 18 instead, RX 5600 XT seems to be the culprit.

Edit: Seems this guy on Reddit has the same issue as mine, replacing GPU solved it for him. I hope it can solve mine. https://www.reddit.com/r/AMDHelp/comments/wdc9kl

 

[emothxughts]

Update: I returned the RX 5600 XT to the original seller under warranty. In the meantime, I put my older RX 580 back in to see if it still works, after I thought it died. Spoiler alert: It still lives, it outputs display, I can play games with it. However, I noticed that in GPU-Z, the bus interface speed only ever goes up to to PCIe x2 3.0.

I also have to mention that I'm running this RX 580 without putting the screws to the back of the PC case, held up the screws for the wifi card right under it. I tried putting the screws back on, the RX 580 will not output a display at all. I unscrewed the card, let it be held by the wifi card screws under it, it outputs display again, plays games with no problems. (Aside from the PCIe bus lane issue).

On the seller's side, he tested the RX 5600 XT with Furmark for like 2.5 hours straight, on a breadboard setup with the card installed vertically, he claimed that it ran with absolutely no black screening or restarts, and no noises either, whereas on my PC it makes creaking & buzzing noises on occasion.

Now I'm thinking the PCIe slot is the culprit. Can a faulty PCIe slot even cause WHEA 18 Cache Hierarchy errors?

Bonus: A video showing what happens when I started my PC with the RX 5600 XT card.

 

[emothxughts]

View attachment 273817View attachment 273818

Update: I returned the RX 5600 XT to the original seller under warranty. In the meantime, I put my older RX 580 back in to see if it still works, after I thought it died. Spoiler alert: It still lives, it outputs display, I can play games with it. However, I noticed that in GPU-Z, the bus interface speed only ever goes up to to PCIe x2 3.0.

I also have to mention that I'm running this RX 580 without putting the screws to the back of the PC case, held up the screws for the wifi card right under it. I tried putting the screws back on, the RX 580 will not output a display at all. I unscrewed the card, let it be held by the wifi card screws under it, it outputs display again, plays games with no problems. (Aside from the PCIe bus lane issue).

On the seller's side, he tested the RX 5600 XT with Furmark for like 2.5 hours straight, on a breadboard setup with the card installed vertically, he claimed that it ran with absolutely no black screening or restarts, and no noises either, whereas on my PC it makes creaking & buzzing noises on occasion.

Now I'm thinking the PCIe slot is the culprit. Can a faulty PCIe slot even cause WHEA 18 Cache Hierarchy errors?

Bonus: A video showing what happens when I started my PC with the RX 5600 XT card.

Update, 11 Dec 2022: Seller tested the card horizontally and tested it in game, he says it black screened. Yes, this is one of the paradoxical situations where a graphics card can pass hours of Furmark stress testing, yet crashes in a game that's likely far less intensive. He also says he'll move forward with warranty.

 

[emothxughts]

Bump

After nearly 2 months of waiting, my RX 5600 XT finally arrives back at the original seller from Gigabyte. They are proceeding to send it back to me.

Now I have a question, what are the chances that a graphics card causes a WHEA 18 Cache Hierarchy error? I currently have a working R5 3600 and RX 580, not a single crash so far, especially not a Cache Hierarchy one. But when I had the RX 5600 XT card, such errors happened quite often, to the point I RMA'd my CPU because I thought it was a CPU problem. The crash still happened with the replacement CPU I got, which got me to think the card was at fault instead, which led me to RMA the card.

 

[Hyderz]

could be a faulty card, ive never had a faulty gpu before (touch wood) but ive seen in tpu forums where gpu crashes in games after a few weeks..
not surprising there..

 

[emothxughts]

Also, when it WHEA 18'd, it's a different APID every time in Event Viewer, which means different processor thread each time...

 

[RJARRRPCGP]

Black screen with backlight still lit=That's a known issue with especially Windows 11, but it won't happen when gaming and Chromium is by far the most likely to cause that issue. I don't know about 10, unless 22H2 is the cause. Possibly a Windows Explorer issue.

 

[emothxughts]

Thing is, I'm on Win 10 and I exclusively use Firefox as my browser. Those black screen crashes occur exclusively in games. Not even stress tests like Heaven and Superposition made the card blackscreen.

Also the black screen mode changes with processors. With the R3 3200G, it buzzes and then black screens with backlights on indefinitely, until I long press the power button. Whereas with the R5 3600, it buzzes and black screens with backlights on, then it restarts after a few seconds.

In any case, the card is on its way back to me. It sure is taking its sweet time though shipping from the west side of the country to the east side where I live. (Look up Malaysia on a map.)

 

[Deleted member 185158]

I have an Xfx RX 6500 XT. Strictly benched it and put it away.

3DMark Vantage seems rough with these 6000 series cards for whatever reason. But I overclocked mine and use Vantage for OC baseline because of how sketchy the benchmark will run.

Push too hard (oc) seemed to lock the system up, accompanying by either a black or Grey screen. I didn't bother looking at the errors, rather lowered the gpu clocks down a bit.

It's a hard lock, generally from just hitting the limit of the chip. This causes degradation without an over-volt. The seller probably just beat the card too much even for the factory OC.

I've also experienced this with an Rx 580, and RX 6650XT.
Most gpus are resilient, but I'm not even one to push on video cards too much. They seem to break easier than processors in the overclocking aspect.

Why Vantage? Cause it gets glitches before the system hangs, more visible than most benchmarks. I figure that runs smooth with my OC, the more modern benchmarks pass with a breeze.

Also, don't beat the card testing with fur-mark. That little program has killed of plenty of cards in the past.

 

[Redwoodz]

I'snt 1.33v awfully high for a stock Ryzen 2 cpu?

Windows update problems on AMD hardware only.. Amazing how often it happens

 

[QuietBob]

@emothxughts
Did you try different display outputs and cables?
Did you try running the games with a 60fps cap and/or underclocked graphics card?
Did you test with a different PSU?

 

[emothxughts]

Update, 4 Feb 2023:
After an agonizing 2 months wait, I finally got my RX 5600 XT card back. Undervolted it, played Genshin Impact for about an hour and a half, nothing but spitting out them frames happens, no buzzing or crackling sounds from the card either, like from before I RMA'd the card. But I have not played another game with it, so I cannot make a conclusion yet.

@emothxughts
Did you try different display outputs and cables?
Did you try running the games with a 60fps cap and/or underclocked graphics card?
Did you test with a different PSU?

I know these questions are kinda moot for a card that was sent out for RMA and came back just today, but I'll answer anyway:

1. I have not yet, I don't have a DP monitor nor cable, but I ordered a DP to HDMI cable, waiting for that to arrive.
2. Yes, and it blackscreened and restarted itself anyway. In fact I remember opening Elden Ring and it crashed before the logos.
3. Unfortunately my other PSU is a 80+ White PSU from AVF, a fairly unknown brand outside my native Malaysia, and it isn't even in the PSU tier list, so I presume it's in the bottom tier. My bad for buying it. My current PSU (the exact make is in the drop down on the left) worked fine with my previous RX 580 (which black screens too, but in a different way, the monitor output merely goes out, everything else is still running and lighting up, check my profile for a thread I started about it.)