PC Graphics

[Toast21]

Ever since I upgraded my gpu from 2080super to 6900xt I get random reboots when gaming. I can be fine for 3 hours and it will reboot the system without warning. At first I thought it was the psu (750w Corsair) but I upgraded that to a 1000w Corsair psu and same problem. Gpu temps stay below 85c and cpu stays below 55c. When I had my 2080 I never experienced such problems. I deleted the nvidia drivers before installing the 6900xt too.

msi z490 gaming edge wifi 

i7 10700k

corsair rm1000 psu

gigabyte 6900xt

32gb Corsair vengeance 

Any help? I don’t know what else to do to trouble shoot this, I get no warning when it reboots, when it is running it runs great. 

[RoboCop]

Well, that didn't work...

Upgrading the PSU to the Corsair HX 1200 Watt still didn't fix the issue... 

I've had it with AMD....!  I'm sourcing a ASUS TUF RTX 3090....!!

I've now tried everything:

Driver only (multiple versions). including (21.3.1, 21.3.2, 21.4.1, 21.5.1)

With SAM (Smart Access Memory), without SAM.

With overclocking, without overclocking.

DCOM Enabled/Disabled.

Re-installed Windows - fresh copy

Updated BIOS to 2201: Update AMD AM4 AGESA V2 PI 1.2.0.2

Specs:

GPU: XFX RX 6900 XT Merc Speedster 319 Black

Mother Board: ASUS ROG Strix B550-F Gaming Wi-Fi (BIOS 2201)

CPU: AMD Ryzen 7 5800x

32GB - 2x16GB Corsair Vengeance Pro RGB 3200MHZ

Samsung 980 Pro M.2 NVME 1TB

Corsair HX 1200 Watt (80+ Platinum) PSU

Corsair H150i AIO Liquid Cooler

3x Gamdias Intake Fans in the Front and 1x Corsair ML120 Pro for the rear with the other 2x Corsair ML120 Pro on the AIO Liquid Cooler

[BattousaiFPS]

• Bro you have my same video card and motherboard. I got news for you...

Go in bios. Go in ai_tweaker...Set cpu voltage manually to 1.35(under vddvr cpu voltage) it's defaulting to 1.47!!!! U have to do this!!!! Also. Do not...I repeat... Do not enable xmp(d.o.c.p) for the ram.  Leave on auto. Also, go in  precision boost overdrive ... disable pfo fmax enhancer. disable precision boost overdrive. Set scalar to manual 2x.set temp throttling to manual 85c. Disable Asus performance enhancement..memory frec to auto... cpu core ratio to 40. Tpu1. Everything else to auto. You may leave resize bar (Sam) on... with that said... In windows.... use amds January (21.1.1) drivers. Do NoT turn on zero rmp. Set fan to 100 percent at 70c or so in the curve, don't change anything else.... do not use any overlays..... turn off steam overlay. Turn off origin overlay. Discord overlay and hardware acceleration too set to off... same with chrome and edge hardware acceleration to off...again! do not use any overlays as the amd software doesn't like it and also do not use its own overlay. No fps or anything overlay ( that's if u want to game)use the game's fps reader if u need it... 

Do all this and you will be able to game. Driver may crash rarely!!!! but this time it will black screen then it will restart driver  itself within windows unlike before... no pc reboot! And u will be able to play for hours without issues. I have tested this now for 3 days...no reboot!!! Try it...

You're welcome

Intel boys: try the same!!! Don't let the motherboard set your clock or voltage automatically. Set it manually accordingly to your chipset...and disable boost so all Cores run at same speed at all times. And do not overclock to its max... only or a couple of hundred Mhz . Mine is 3.8 I got it running at 4.0ghz no issues... when I had it overclocked to 4.6 and boost was on, there was lots of issues with amd driver for gpu and this settings. And some games wont even run...(like black ops cold war... try it... if u enable xmp u won't be able to play coldwar)

Let me know how it works for you boys.

[RoboCop]

@BattousaiFPS 
Thanks for the post, however I followed everything to the tee, but got another random reboot after about 2h of gaming in COD MW/WarZone, which is the only game I play for lengthy periods.

I went into the BIOS and first reset all previous changes to defaults - F5. Then proceeded setting up the settings mentioned.  Please check if I have not missed anything... 

I then set TPU to TPU1 and rebooted in order to manually set the CPU Core Ratio to 40 for 4Ghz

I then switched on SAM ensuring that CSM is Disabled.

Then Saved all Changes made in the BIOS.

I then used Adrenaline driver 21.3.1.  Will load the 21.1.1 driver tomorrow again.  I then also disabled all Hardware acceleration on Edge Chromium and Discord, Steam and any software that starts normally. Only have LGHub and Eset Anti-Virus.

I then reset the Adenaline Software to Automatic - Default, with a custom curve of 100% above 70 degrees Celsius.  I also set Zero-RPM to disabled.  I switched off all startup programs and did not have any of the following open: MSI Afterburner, HWInfo, TechPowerUp, everything.  No Origin, Epic or UBiSoft is running and no Riva Statistics tuner either, so I was flying completely blind here...

After about 2 hours I got a reboot.  Checked the Event Viewer and only got a couple of DCOM warnings about permissions and then the Error on Kernel-41 that said it experienced an unexpectedly power failure with boot error code 0.

As you can imagine I’m fed up at this point so managed to get the very last RTX 3090 in the country, literally . Paid double the price of the RX 6900 XT as I’m so fed up on this issue.

[hunter005]

I'm having a suimilar issue with 6800XT. Could you share what you got in the last crash (minidump), please?

[RoboCop]

@hunter005 

That is the thing I don't have any crash dumps created...  Only a restart, as if the power was cut the the Power Supply.

[BattousaiFPS]

@hunter005 , your PSU looks a bit too close to the minimum requirements for that card... I would suggest at least  a 1000w PSU... OR 850w minimum. Give my post a shot (scroll a few post up...)try the settings that I suggested RoboCop to try, on your setup. You may get lucky like me and be able to play 80 percent of the games like I am.... The issue is clearly a combination of driver crashes, adrenaline and cpu settings as this is what I have found after troubleshooting for a long time now and thanks to that I am able to play many aaa tittles without crashing... (not apex). Let me know if any of these changes helps for you.

Make sure you are using driver 21.1.1!

What I do want to hear is from any intel users trying to disable boost as well to see if issue changes with intel cpus as well....

[hunter005]

Guys, I made somethings that I think solved my problem, and I would share it, after 2 days playing Warzone without any crash:

1- I changed the windows page file from "Automatically manage paging file size for all drivers" to "custom size: Initial 20000, Maximum 30000

2- Uninstalled and removed with DDU, old driver, and installed the version 21.1.1.

3- Modified the adv_Options file, of cod Warzone (C:\Users\*YOURUSERNAME*\Documents\Call of Duty Modern Warfare\players\adv_options.ini) with these values:

VideoMemoryScale = 0.6

RenderedWorkerCount = 4 (my cpu has 6 cores, but I wanted try the default value)

After that, didn't crash any more here.

I hope I helped you. Let me know if something work for you guys...

[RoboCop]

@hunter005 

I think your solution could definitely also work, mainly due to the fact that you are probably limiting the 8 Core processor (if you are running the 5800x) to only use 4 cores - RenderedWorker ount.  As from YouTube Videos I saw that this setting should be equal to the number of cores in your processor.  I also previously had the VideoMemoryScale set to 0.55 at some point, but I think the setting that perhaps has the biggest impact is the RenderedWorkerCount.

What hardware are you running?

[hunter005]

My Specs:

Gigabyte 6800XT

Asrock AB 350m Pro4 (Last Bios Update)

Ryzen 5 3600

2 x 8GB DRAM

Corsair CX 650W Bronze

1xHDD, 1x SSD, 1 NVme

I'll try to remove the modifications that I've done, one by one, to try to discover what really solved the problem. I think the parameter "Renderworkercont" didn't had impact performance here. I'm not 100% sure, I will test this week, because its hard to isolate the things, since the crash takes about 2 or 3 hours to happen...

[Lego_Underfoot]

For Apex legends I found that turning on triple buffered V-sync stops my graphics card from restarting my computer. Which is annoying cause it locks my frames at 60 when I have a 144hz monitor, but at least I can play games. I really wanna know what is happening that is causing the gpu to restart the PC.

[RoboCop]

Is this then not perhaps related to the Enhanced Sync portion of the Adrenaline software AMD Releases New Radeon 20.2.2 Drivers to Fix 'Black Screen' Bug - ExtremeTech

[BattousaiFPS]

@RoboCop 

NO, that is a different issue. This drivers they are putting out are FULL OF ISSUES that they are not fixing.

[RoboCop]

I'll definitely look at upgrading the PSU to at least a 850Watt PSU.  Maybe yours is related to a Power issue.  Other than that perhaps the overall consumption is less than the max when only using 4 cores of the CPU.  When pushed to 7 or 8 cores it might be consuming more power than is available...  Do you get a beep or error lights from the Mobo (if equiped)?  As this is what I'm getting.  A reboot, with the VGA light on the Mobo being lit.

[hunter005]

What about the 1200 Watt PSU you tested before? You said that not resolved your issue...

Did you tried change page file size, as I suggested?

This kind of issues (long time to occur) looks something related with memory leak... We know warzone is a very bad coded game, having a lot of bugs, glitches, etc.

I really spent too much time trying to tweak this game to avoid the frames sttuterings.

[vincenzo34]

Hello,

I had the same problem as you after purchasing an MSI 6900XT GAMING X TRIO card. Namely, black screen, then reboot of the PC in play (especially WARZONE) after a certain time (1h ... sometimes more).

My config:
10900K (No OC), 32GB, MSI Z490 Unify, Be quiet 1200w platinium.

I then tried everything starting with deleting the XMP profile, increasing VCCIO & VCCSA, lowering the card voltage, the frequency, limiting the FPS, running the fans at 100% ... etc. finally all the "solutions" possible ... but nothing or did something, always that **bleep** black screen then random reboot in the middle of Warzone ...

Being convinced that the problem indeed came from the graphics card ... (hardware or software drivers) because Proc and ram tested on OCCT and MemTest. I succeeded in ordering another 6900xt, the Sapphire 6900xt nitro+ SE (for the same price as MSI). I just plugged it and since NO MORE ISSUES after 10 hours of Warzone, Cyberpunk .... The drivers being set to Default ...

To summarize, the MSI 6900XT GAMING X TRIO that I received had a problem ... (Hardware, BIOS ... I don't know what ...) so I would advise buyers of this card to be wary of its stability...

If that can help someone ...

ps: sorry for my english, i m french ...

[hunter005]

@vincenzo34 

Could you share the parameters you got from minidump file, if you still with it?

[vincenzo34]

sorry but no dumpfile...

[hunter005]

I modified the parameter Renderworkercont to 6, again (I have a six core processor), and still good, without crashes, after 4 hours playing warzone (before, the crashes happened after 2 hours always)

I would remenber you that still using VideoMemoryScale 0.6 and page file size in 20000-30000.

Next days I will test reset one by one the values I changed, to default values again, to discover what's causing the issue, and will share with you.

[vincenzo34]

Okay, I come back to the news ... which is not very good ...
I again had a crash after 4 hours of Warzone ... with my new Sapphire 6900xt nitro + SE.
That's enough for me, I go back to my Titan X pascal and will wait for a 3090 or 3080ti.
Amd it's over, too wasted time.

[RoboCop]

@vincenzo34 

In a way that is good to hear, because I'm also tired of this instability issues with the RX6900 XT.  If you had the same issue with 2 different AIB models then it can only be a Driver issue or the Card itself.

Hopefully I'm getting my new NVidia RTX 3090 today and I will confirm if the issue is finally resolved. 

[RoboCop]

@hunter005 

I got another reboot yesterday.  This was after the reverting the BIOS to 2006 of the motherboard.  I had the Standard preset on the Adrenaline software.

I will try a couple of things today (still on the old 20.12.1 driver). 

I will configure Windows Page Size Settings to min: 20000Mb and max to 30000Mb

For COD MW/WarZone (adv_setting.ini) - Set the RenderWorkerCount to 4 and VideoMemoryScale to 0.6

I will also disable the Enhanced Sync in the Adrenaline Software.

I have e-mailed AMD directly - no response thus far

I have e-mailed ASUS directly - which has come back saying I should remove the CMOS battery for 30 seconds after flashing to Version 2006 and then replace it again, attempt to switch on with power cord unplugged and then try again.  I'm sure this will update the BIOS, but doubt that this is a motherboard issue in anycase.

I have also logged a support ticket with XFX directly - no response thus far

After all these hours and hours of troubleshooting, trying to find the problem (no regular consumer would go through all this trouble) in anyways.

I'm pretty sure, as soon as I get my new RTX 3090 all my problems with reboots will suddenly evaporate in the air and just be a memory of the past.

Now to get some other poor "smuck" to buy a second hand (3 months old) XFX RX 6900 XT, because I really doubt I will get my money back...

[RoboCop]

@Matt_AMD 

Yip, I got a reboot and then white light (vga light) on motherboard lit.

I've tried both single rail mode as well as multi-rail mode on the Corsair HX 1200 Watt PSU which has a 70a limit on single and 40a on multi-rail mode.

Both 8-pin connectors are coming straight from the PSU (not daisy chained). 

I agree with @BattousaiFPS something (software related - The Driver itself) is causing all RX 6900 XT cards to either go into extreme temps or drawing too much current or over faulting or something.

This only happens during lengthy gaming sessions where both the GPU and CPU work together in order to render images. 

So from heavy loads to light loads and this causes a trigger of some sorts.

I've never had a reboot during benchmarking in 3D Mark's TimeSpy.

The titles that I've experienced the issue in include:

Call of Duty Modern Warfare 2019

Shadow of the Tomb Raider

Apex Legends

[Matt_AMD]

I reviewed your post and system specs on page 7. We are actually using the same XFX 6900 XT Merc, though mine is the Ultra Edition. Our specs are broadly similar, however i have a 5950X and a HX1000w platinum PSU. 

I see that you have been troubleshooting for some time and have tried many different system, OS and game tweaks so i can understand your frustration. 

I do not experience any system reboots with the 6900 XT, so it may be that in your case the GPU might be faulty. You should be able to validate this if the 3090 works without issue, and then you can proceed with an RMA with your place of purchase for the 6900 XT. 

I will also add that i use the motherboard BIOS default settings. Nothing is changed in terms of voltage, overclocking etc. Reviewing the discussion i can see that many people are adjusting BIOS settings, overclocking, using PBO etc. This can all lead to potential system instability so when trying to troubleshoot stability issues i would recommend putting everything back to the default BIOS settings. 

We have a kb article that users may find helpful to troubleshoot system stability issues. 

Troubleshooting Tips for Resolving System Stability Issues | AMD

[Lego_Underfoot]

In addition to @Matt_AMD, I have my sapphire 6900xt going out as an RMA, just sent it out yesterday, and a new motherboard coming in today. When I get my graphics card back I'll wipe my windows drive and start from scratch and see if that fixes it. 

[Matt_AMD]

Thanks, please let us know how it goes once it is replaced. 

[hunter005]

@RoboCop 
Did you tried to put your Computer on heavy load, without gaming? Using softwares like Prime95 (to Stress CPU) and OCCT GPU (to stress GPU), at the same time, to try to make a max drainage possible, and verify if your PC reboots? In this circunstances, my PC didn't restart... This kind of stress is most than a game usually cause...

I'm Really thinking what solved for me was the parameter VideoMemoryScale, since now I'm using RederedWorkers at my maximum (6), and the recurrent error I've got, registred in to minidump, 0x0000010e is related, by Microsoft, with "VIDEO_MEMORY_MANAGEMENT_INTERNAL"

[Lego_Underfoot]

@hunter005 

I seem to have the same issues as @RoboCop , I ran furmark and cinebench. Cinebench runs for half an hour without an issue, I can't run furmark on you more than like 7 minutes before my computer restarts.

[RoboCop]

@hunter005 

I did a OCCT last night for 30min-1h each and no crashes.

[Muppet]

Get a better power supply like really decent one not cheap one, wattage is not the only thing that is important.

[Enohriel]

You guys are lucky. You can at least launch games.
My driver is crashing as soon as I try to launch a game.
I think my brand new 6900xt is just dead.
Ive tried a lot of fix without success..

[Danoisyone]

You tried the disable com?  That's what fixed it for me, but I had already fresh installed windows and tried several driver versions.

[Enohriel]

Yes I did, same issue.

[Muppet]

think i tripped ocp or something this morning booting into windows altho my system is very heavy like 5 ssd 1 nvme 2 mechanical drives and my old gtx 1080 is also still powered next to my 6900 xt cos i am waiting for waterblock to arive for my 6900 xt cos i want it to last 10 years not 5 years, altho i probably upgrade in 3 to 4 years when chip shortage is resolved.

system would not turn on until i remove pci-e power cable from my gtx 1080 and reseated power, hard to get a 1000+ watt power supply these days

[mc_line]

I will take aware my previous comment. Upgraded from 850 watts to a 1000 watts power supply fixed my random gpu crashes. Still AMD drivers are buggy as hell.

[Muppet]

those power spikes can hopefully be adressed in drivers, anyway im getting my waterblock for my 6900 xt next week thank god void warranty sticker is invalid in my country altho i would expect not to ever need warranty as gpu's running cooler last quite a bit longer then the warranty last.

[eliotcole]

Upfront

Just making my first post here, to add to the amount of users experiencing this issue. I'll list full stats and all that when I can.

Also wanted to say:

I'm determined to stick with AMD on this, as over the past decade NVIDIA's quality of life usage (everyday stuff) is just horrific, and their support forum is literally only useful for the occasional nice nerd.

I know there will be issues with any new hardware, especially bleeding edge stuff like this. Plus, having worked tech support my entire life, I appreciate how difficult some firms feel it is to address Problems, and the many Incidents that collate to indicate a given Problem.

I hope that in reading this (and the many other threads about this issue just on these forums) that AMD are able to see all of these technically separate incidents and see that there is a problem behind them all.

I hope that my using ITIL language has helped in that, @Matt_AMD .

Still, a frank mention that the power spikes are an issue could really help the community. More importantly, if it's only stated here, then AMD don't run the risk of it becoming a 'big thing'. Since I'm aware that you're probably beavering away on a fix for this, believing it's better to work 100% on the fix and not mention the fact that it *needs* fixing. I can tell you from a customer service perspective, it's better to just let us know you're working on a fix for it, and perhaps even provide some way people can help with test results, etc.

In leiu of a fix, an official work-around would be ideal. Again, all ITIL language intentional. 🤭

Right, back to the issue ...

Discussion

The conversation around the DCOM has made me wary about changing it, because I make a lot of changes to my system to stop various Windows 10 services sending data elsewhere, I don't wish to undo any of that hard work by mistake.

However, purely thinking laterally about it, is there a way to needle the exact DCOM process/service/device, and alter the settings for that to ensure the same outcome is received without banjaxxing everything?

I can't find my most recent crash but I'd imagine if we can all identify the DCOM errors that lead into the powerdown and either a single or a few items to modify, we could have a patch on our hands. We could also (likely) write a Power Shell script to make the changes safely, and to reverse them if/when AMD have fixed this.

---

EDIT - I am making a guess that this was the server that caused the DCOM issue:

354FF91B-5E49-4BDC-A8E6-1CB6C6877182

According to the Component Services window that corresponds to:

C:\Windows\System32\SpatialAudioLicenseSrv.exe SpatialAudioLicenseServerInteractiveUser

Now, this is odd, because that's an audio component, but it's given NN criticals before what looks like a shutdown I had.

---

Setup

Here's my basic setup:

Category	Detail	Notes
Operating System:	Windows 10 Pro
Motherboard:	Asus ROG Strix X570-E	Latest BIOS - Updated by Scan 3XS
CPU / Processor:	Ryzen 5900X
GPU:	Powercolor RX 6900 XT	OEM
Memory:	2 x 16G Corsair Vengeance LPX DDR4 3200
Power Supply / PSU:	Corsair HX1000W

Crashes

Incidentally, I have seen the issues in:

• Total War: Warhammer 2 - When Grom wins a siege battle at a nearby settlement.
• 3DMark Time Spy benchmark - Just when the avatar turns back to explore the museum after the second giant walks by the gaping hole in the wall.

I have not attempted to try the Total War: Warhammer 2 trigger in DX11, and am wondering if Direct X 12 is a component of the overall problem that is causing this incident.

Conclusion

Like I say, I'll try to update this when I have more, plus I'll add that Scan are actively pursuing an RMA on the card. If there is actually no physical reason for this, I find that to be overkill, plus in this market, what would I do in the meantime? Move back to a single 970 (all I have left of SLI, as I gave away my other one to a relative) or use the onboard? I think not.

Especially when this could definitely *BE* the fault of the Graphics Card, however there might be hardware issues elsewhere (yes, on all the various systems). As in, there could be a capacitor on all the motherboards that isn't able to take the spike that the card is requesting.

Equally, we could all be using single CPU power cable here, when it should be 8-PIN and 4-PIN. I dunno, I'm just saying it'd be good to find a way for us to all pool the data and find some common threads, here.

I'd also like to identify the individual component to disable DCOM on, too.

[RoboCop]

A Further update from my side...

So I did manage to receive my NVidia ASUS TUF OC RTX 3090, removed the XFX RX 6900 XT Merc 319 Black and replaced it with RX 3090 card, loaded latest drivers from NVidia and was up and running within 15 minutes.   No problem.

I then received a response from AMD whereby I needed to provide them with the results from DirectX - dxdiag results, so I removed the RTX 3090 again and replaced it with the RX 6900 XT.  I then did the following actions as my last attempt in order to get rid of the random reboot...

• I had reverted back to non-Beta version 2006 of my ASUS ROG Strix B550-F Gaming Wifi motherboard.  But never removed my CMOS battery.  So I did this, waited 30 seconds whilst the power was disconnected, attempted to turn the power on to drain reserve power in the MB and then inserted the CMOS battery again.
• I then loaded BIOS defaults (F5), enabled Resizable Bar support (SAM) and then enabled D.O.C.P again for the memory running  @3200Mhz.
• I disabled DCOM settings in Windows as previously mentioned as this is not needed on my Gaming PC
• I configured Windows Page Size Settings to Min: 20000Mb and Max to 30000Mb on my 32Gb system

• For COD MW/WarZone under My Documents\COD Modern Warefare\(adv_setting.ini) - I then set the RenderWorkerCount to 4 and VideoMemoryScale to 0.6
• I then proceeded to unload all drivers and Sound Card software using DDU (Display Driver Uninstall Tool)
• I then installed only the 21.4.1 "Minimal Install" i.o.w, no Adrenaline Software, only the driver
• Rebooted and finished any other installations.

I was able to play the following games without rebooting whatsover...

Apex Legends - over 3 hours

COD MW/WarZone - over 4 hours

COD Black Ops Cold War - over 5 hours

F1 2020 - over 2 hours

• Was it perhaps the card that was not seated properly in the PCI-E slot?
• Was it the BIOS that needed a CMOS battery clear?
• Was it the DCOM Setting?
• Was it the Windows Page File Setting?
• Was it the 21.4.1 driver only?
• Or is this still purely speculation...

Anyways, after my dxdiag results were given to AMD Tech support, the only thing that they responded was to remove the driver again and load the 21.5.2 driver.

I've given up on the AMD card, will probably sell it soon.  Have not had a single problem with the RTX 3090

So I had uninstalled all drivers again using DDU

[Lego_Underfoot]

Someone mentioned previously that changing the dcom settings fixed there issue. I just feel like that kind of a BS fix. It could be the PCI-E slot. I know my gen4 x16 slot was broken on my motherboard so I had to move it to the gen4 x8 slot. I only seemed to get the issues when I was in the x8 slot.i can't say that for sure cause it could have been a driver update as well.

[eliotcole]

Just trying to parse this (apologies, I can be stupid sometimes) ... it seems like you got the 6900 working faultlessly, and then decided that it took too much/long to get to this point, and thus are sticking with the 3090. Or was all of that testing on 3090?

Sorry, just being dumb!

---

@RoboCop wrote:

A Further update from my side...

So I did manage to receive my NVidia ASUS TUF OC RTX 3090, removed the XFX RX 6900 XT Merc 319 Black and replaced it with RX 3090 card, loaded latest drivers from NVidia and was up and running within 15 minutes.   No problem.

I then received a response from AMD whereby I needed to provide them with the results from DirectX - dxdiag results, so I removed the RTX 3090 again and replaced it with the RX 6900 XT.  I then did the following actions as my last attempt in order to get rid of the random reboot...

• I had reverted back to non-Beta version 2006 of my ASUS ROG Strix B550-F Gaming Wifi motherboard.  But never removed my CMOS battery.  So I did this, waited 30 seconds whilst the power was disconnected, attempted to turn the power on to drain reserve power in the MB and then inserted the CMOS battery again.
• I then loaded BIOS defaults (F5), enabled Resizable Bar support (SAM) and then enabled D.O.C.P again for the memory running  @3200Mhz.
• I disabled DCOM settings in Windows as previously mentioned as this is not needed on my Gaming PC
• I configured Windows Page Size Settings to Min: 20000Mb and Max to 30000Mb on my 32Gb system

• For COD MW/WarZone under My Documents\COD Modern Warefare\(adv_setting.ini) - I then set the RenderWorkerCount to 4 and VideoMemoryScale to 0.6
• I then proceeded to unload all drivers and Sound Card software using DDU (Display Driver Uninstall Tool)
• I then installed only the 21.4.1 "Minimal Install" i.o.w, no Adrenaline Software, only the driver
• Rebooted and finished any other installations.

I was able to play the following games without rebooting whatsover...

Apex Legends - over 3 hours

COD MW/WarZone - over 4 hours

COD Black Ops Cold War - over 5 hours

F1 2020 - over 2 hours

• Was it perhaps the card that was not seated properly in the PCI-E slot?
• Was it the BIOS that needed a CMOS battery clear?
• Was it the DCOM Setting?
• Was it the Windows Page File Setting?
• Was it the 21.4.1 driver only?
• Or is this still purely speculation...

Anyways, after my dxdiag results were given to AMD Tech support, the only thing that they responded was to remove the driver again and load the 21.5.2 driver.

I've given up on the AMD card, will probably sell it soon.  Have not had a single problem with the RTX 3090

So I had uninstalled all drivers again using DDU

---

-------------------------

---

Also, has anyone tried Linux?

[RoboCop]

@eliotcole all this testing and troubleshooting was done for the AMD RX 6900 XT and yes I managed to work faultlessly for the duration of the testing.  However you know Murphy and it might just be a placebo effect or inadvertently it could have been resolved for good, I would not know.  So my questions are still left unanswered...

• Was it perhaps the card that was not seated properly in the PCI-E slot?
• Was it the BIOS that needed a CMOS battery clear?
• Was it the DCOM Setting?
• Was it the Windows Page File Setting?
• Was it the 21.4.1 driver only?

I’m now running the NVidia RTX 3090 and have no issues, even from the start...