PC Drivers & Software

[0x87]

System specs

GPU: MSI Radeon RX 6700 XT Mech 2X OC 12GB
CPU: AMD Ryzen 9 5900X Processor
Motherboard: MSI MAG X570S Tomahawk Max Wi-Fi Motherboard
PSU: Thermaltake Toughpower GF1 ARGB Gold 850W Power Supply
RAM: Corsair Vengeance LPX 32GB (2x16GB) 3200MHz CL16 DDR4
HDD: Samsung 970 EVO Plus NVMe SSD 1TB
Monitors: Samsung 27" Flat Odyssey G50 & Samsung Odyssey G5 QHD 144Hz FreeSync Curved 32"
Windows: Windows 10 & 11 issue persisted
Driver: Adrenalin 22.11.2 & Adrenalin 22.5.1 Issue persisted

Issue:
For months I have had zero issues but last week while playing a game I had my game freeze, screens go back, music continued to play and act as if the PC was running normally, I tried everything to exit the game and restore screen to normal, I needed to do a hard shut-down, when turning PC back on only one screen would work, and drivers were disabled, requiring me to re-enable them via device manager, this issue continued to occur on other games, I have done a clean install of windows 10, new drivers same issue, clean install of 10 with older drivers, same issue, clean install of windows 11, both drivers (DDU clean up before installing newer drivers) same issue, tried using only 1 monitor, same issue, tried replacing cables.. same issue

Can anyone help me.. or do I need to RMA my card?

[ThreeDee]

I ran that GPU .. great card .. it's now in my daughter's setup and running great.

How old is  your PSU?

Are you running separate cables from your PSU to each power input on your GPU?

Are you running the latest AM4 Chipset drivers from AMD.com?

Is  your BIOS up to date? SAM working?

CPU/GPU temps ok?

[0x87]

The PSU along with the whole system was built on Sep 1st, I'm 99% sure the cables are separate but I will confirm in 10 hours after work, bios was updated by store I purchased my MB from, I am not sure about the SAM but will check, I'll do a fresh install of windows tonight and install newest chipset drivers, any recommendations on what drivers I should use for the GPU?

CPU and GPU temps appear to be normal 

Furmark for 30min has my gpu at less then 100c, I can't remember off the top of my head but I think it was at 80c

CPU is cooled via Cooler Master MasterLiquid ML360L ARGB 360mm AIO Cooler

The issue itself has only arisen in the past week, from September 1st till last week I have had zero issues running games like cyberpunk 2077 and etc, yet now games like "we who are about to die" or "drug dealer simulator" crash after about a hour, 

The sound will still run, screen will freeze, I lose signal in Monitors and require hard reset, upon reset need to re-enable drivers in device manager.. 

[ThreeDee]

anything get installed/happen .. or similar this last week prior to issue starting?

What BIOS version are you currently running? 7D54v14 is the latest which came out right before you purchased your PC

As far as GPU drivers .. I always run the latest "greatest" without issue on my daughter's 6700xt .. my 6800 .. and my wife's 5700xt .. some seem to have issues with certain drivers of which I have yet to encounter so I got nothing for you on that one unfortunately

IF your PC got jossled around (bumped) prior to issue's happening .. or moved .. maybe simply just reseat everything and see if that does anything for you.

Check your system's Event Viewer for possible logged errors that might maybe give more insight to your issue

[0x87]

I will check my bios version when I am home and double check the connections, I will also mention I'm using a pcie extension cable as my GFX card is mounted on a wing to my case

Event viewer only shows critical from hard shutdowns.. I'll post the errors when i get home but I don't recall any before the crash

Nothing to my knowledge has been installed on my PC, considering I'm doing complete fresh installs, installing drivers and then a game to test the issue,

Only software on pc at the time of crashes are drivers, steam, discord, game hub, malware bytes, MSI centre, Chrome and obviously the game I'm testing 

[ThreeDee]

I'd take a gander at your PCIe riser cable too .. and any other cable extensions you might have inside your PC

plug your GPU directly into your motherboard if you can and spin up some test games

[0x87]

Wouldn't a riser issue have more visible issues? My issue only tends to show up after a hour or so of gaming, that being said I'll have a look at it..

[drewster]

As ThreeDee said earlier, make sure all connections are good.  Also make sure you have 2 separate PCI cables going to the GPU.  Also want to know if there is any overclocking being done.  Can also try installing AMD chipset drivers 1st, then all other drivers, then the GPU drivers. Try not installing any other program (discord, game hub, malwarebytes and the MSI center), just what's needed to get the game going.

[0x87]

There is no over clocking being done, I haven't touched any of those settings just pretty much plug and play when I built this pc, I will double check all cables when I'm home in 7 hours

I'll reinstall windows 11, only install drivers and steam, see how we go from there.

[0x87]

I would like to add this issue started on the 8th of Dec.. I was playing games with zero issues prior to that date, using Windows 10 and latest amd drivers

Last game I was playing with no issues was SCUM

[drewster]

Hmmm.  Were you still using Win10 when the issues started, or Win 11?

[0x87]

Windows 10, I upgraded after issue began 

[macchky · Accepted solution]

Try to disable MPO(Multiplane Overlay)
https://www.techpowerup.com/forums/threads/feature-request-disable-multiplane-overlay-mpo.289163/

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\Dwm]
"OverlayTestMode"=dword:00000005

The lack of this key means its enabled

One Click reg file (mpo_disable.reg) is here

https://nvidia.custhelp.com/app/answers/detail/a_id/5157

[0x87]

I'll add this to the list of things to try when I get home in 2 hours 

[0x87]

I have gone 3 days without a crash, after disabling MPO, but that said I'm also not playing the games that caused the crash

I also have installed drivers only WITHOUT Adrenalin 

If it crashes again I'll update 

Thanks for this advice 

[nightmare13255]

Helped fix my issue. Thanks so much. Hasn't happened since.

[MADZyren]

I think I would try to set a limit to FPS to reduce GPU's power draw to see if that changes how it behaves. While you have enough watts, Thermaltake is not the best possible brand. By reducing framerate and thus power draw, if problem goes away, there is a good change it is either GPU or PSU, but not drivers/software issue. If problem does go away, I would consider trying with another PSU. Reduce fps a lot to make a real difference. Set limit to like 60 where it would normally do 80-100 and so on. Drop at least like 20% of fps.

[Liabelz]

Same issue.

I’m running the same GPU RX6700XT. The funny thing it can handle furmark stress test about 3 hours with 77 degree temps max. No lagging, no black screen, or stuttering.

[0x87]

Turns out I was using 1 cable that splits into 2, I'm now running 2 of those cables, about to reinstall windows 11 and will update with findings 

[KabouterPlop1]

My card can handle 24 hours heaven followed by 8 hours port royal and 8 hours superposition and will have many gpu driver crashes and freezes even system freezes with audio still playing in world of warcraft.

Appearntly AMD just sabotaged their rma page to be down until next year as well probably cos they get a ton of rma request due blackscreens on their new gpu's, those cant even rollback to more stable drivers because there aren't any.

[MADZyren]

If computer used to work with this hardware, bios-version and so on, it is not likely to be because of those.

Things that might have changed:
- Windows update has updated something
- Game files were corrupted (Steam and Epic atleast have option to check them)
- Problem with shader cache (try clearing shader caches)
- Some file got corrupted (run chkdsk /F and sfc /scannow) in command prompt as administrator)
- Could be a memory issue. Those are sometimes very difficult to diagnose, but try disabling XMP from BIOS to diagnose.

EDIT: I did not notice you mentioning it is stable with Furmark, so removed PSU from list.

[0x87]

That's why I'm at a loss.. I was playing darktide, cyberpunk and etc nil issues and over night this has arisen..

Wouldn't the issues such as corruption of files be erased by reformats?

If I diabale xmp what am I looking for to confirm this was the issue?

[0x87]

It looks like A-XMP was already disabled, should I enable it?

[MADZyren]

Yes, but it will hardly fix the issue.

[MADZyren]

Reformat/reinstall would fix corrupted files IF you also removed game files (which can be on a different hard drive than Windows. Of course system files will be fixed by reinstall of C-drive alone)

Disabling XMP would tell if there is an issue with memory settings. My previous CPU could run very tight subtimings when new, but fairly soon I had to loosen them to make system stable again. I have never encountered this before, but seems like it can happen. Also if memory is "almost" stable, it can seem to run normally and appear to be ok, but cause random issues. In my case, games worked, but Prime95 and Folding@Home gave errors. If disabling XMP works, then you can re-enable it, maybe increase voltage to say 1.375V, maybe drop clockspeed one step down or something like that. If it's memory controller that can no longer maintain that speed, that dropping clockspeed a bit or increasing latency settings would fix it.

Since GPU has been tested with furmark, you could also try dropping CPU clockspeed to see if it can't handle the strain or maybe use some CPU intensive benchmark and run it time and time again. Maybe for instance 3DMark CPU benchmark or Cinebench to see if they crash.

[0x87]

I am now aware the date my issues started is the date the new AMD drivers we're released, I'm going to do a fresh install of the previous recommended drivers and see if I can block windows update 

[0x87]

Fresh install of windows 11, newest drivers installed, 10 minutes into gameplay and hard reset required, drivers disabled again

Nothing has worked 

[MikeySCT]

Well, got two things on my mind concerning your issue:

1. Are RAM in the proper slots? Usually, they should be in slots 2 and 4, not 1 and 2, or 1 and 3...

2. Are you sure RAM sticks, and GPU are properly and securely seated in their slots?

I hope that you don't take offense, not implying you're a noob or that you dunno what you're doing, sometimes, the issue that can rear its ugly head can be downright puzzling.

[0x87]

They are in 2, 4 and GPU is using a riser.. since enabling XMP profile 1 I haven't had a crash.. it's been only an hour though 

[Liabelz]

Try to disable AMD Crash Defender, it work for me at least for now

[ExElShredder]

You guys can troubleshoot all you want, AMD is just worthless Trash if we talk GPU. Playing NFS Unbound with the newest drivers is impossible because it keeps stuttering and eventually crashing, if i revert to an OLDER driver that DOES NOT have support for Unbound i get stuttering after playing around 1 Hour before is freezes and crashes. Already tried ALL the **bleep**, XMP Profiles for my RAM, Older Windows updates, newer Windows updates, unplugging certain Periphery, Reinstalling Windows 10, diffrent PSU; diffrent RAM, diffrent Mainboard and many more of the usual **bleep**ty ways to fix this crappy overpriced GPU (6900XT), wont help and i cant get a proper refunde because i would need to wait 8!!!! Weeks to get any feedback. These Forums here are a joke too, the Bug reporting in adrenaline is useless too, you never get any **bleep**ing feedback what might have happened and appearently this bug exists since the VEGA 56 GPUs, ill go to my **bleep**ing lawyer tomorrow (got insurance) and make sure i get my money back. Cheers AMD you **bleep**s.

[0x87]

This has nothing to do with my issue...

[0x87]

I was able to play a game for 50 minutes today before crashing, disbaled drivers once again 

[KabouterPlop1]

Try 22.3.1 drivers 22.5.1 aint stable either search for drivers on guru3d since AMD decided to stop linking it in previous driver visions, i am having system freezes while playing world of warcraft so far 22.3.1 has been fully stable, i do not trust 22.5.1 either.

Wow is especially broken by it self as well cos if the gpu driver crashes it freezes the whole system so you wont have a tdr and ability to do a bug report, deadlock seems to effect nvidia users possibly intel cpu users as well as game has a conflict with 1 of its cpu threads.

on 22.11.2 for some reason it uses totally different threads with 1 cpu thread maxing out 100% while on 22.3.1 it does not do this.

If had no system freezes yet for almost a week now

[0x87]

After spending 3 days troubleshooting I have taken to my local shop to get the issue diagnosed so I can find out what the issue is, at this point its looking like a Ram issue

I will update once issue is diagnosed 

[Arnt]

Did you try to disable the MPO like one wrote earlyer? i did have a ton of issues with my system since may this year, and all was Related to AMD gpu driver, but disabling the MPO fixed ALL of my problems.

[0x87]

To be honest I forgot about this 

[223hpyhooker]

Hopefully the issue gets resolved and you can let us know what the issue was.

[nightmare13255]

Having the same issue, however its not disabling my drivers at all. When I hard-reboot, my drivers are fine. I don't receive any errors, just acts like the pc just turned on yet again for the first time.

[farmeunit]

Not quite the same issue, but had a friend with a Vega56.  Built a new machine but games kept crashing and locking up.  Had a brand new 750 Corsair PSU.  I couldn't find anything wrong, so told him to swap power supply.  Tried an RM850 or maybe more.  Same issue.  Tried a Thermaltake and had zero issues after that.  I was a big fan of Corsair and hadn't seen that particular issue before.  I have changed to Seasonic, but sometimes things that appear unrelated, are not.  It's always good to try removing a stick of RAM at a time, etc..  Usually you'll get BSoDs and such with that, though.