WHEA_UNCORRECTABLE BSOD - Now FPS Drop Problem - RX 6700XT - R5 5600

[000me]

System
GPU : Sapphire Pulse RX 6700XT
Motherboard : Gigabyte B550M DS3H rev1.4
CPU : AMD Ryzen5 5600
Ram : 3200mhz 8x8gb
PSU : 650W

Hello,

I bought a Sapphire Pulse RX 6700XT graphics card. When I tried for the Red Dead Redemption 2 game for the first time since I got it, I started getting errors after about 10-20 minutes. My CPU Temperature is around 80 degrees, my Graphics Card Temperature is around 72 degrees. After 10-20 minutes I get WHEA_UNCORRECTABLE error.

I tested the ram with Memtest, there is no problem. I did a CPU test with Cinebench, no problem, I did a GPU test with FurMark for about 30 minutes, I did a GPU test with OCCT, no problem. According to the tests you can see, there is no problem. On Reddit, it was said that if I changed the PBO and CBS features of the processor from Auto to Disable, the problem would be solved, I did that, but the result did not change. I would be very grateful if you could help.

Also I wanted to try it in Witcher 3 game, maybe because it was in the game, I set it to Ultra+ settings, I played the game for about 20 minutes and this time the game crashed (it gave a crash error), not the blue screen. Average GPU temperature in-game is 70-72 degrees, CPU temperature is around 80-83 degrees. But a CPU saw 89 degrees.

After that i analysis my minidump file and i found processor is failing here. I'm also sharing my minidump file from WHEA Uncorrectable bsod error.

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
nt!_WHEA_ERROR_RECORD structure that describes the error condition. Try !errrec Address of the nt!_WHEA_ERROR_RECORD structure to get more details.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: ffff8b8cdbe7b028, Address of the nt!_WHEA_ERROR_RECORD structure.
Arg3: 00000000bc000800, High order 32-bits of the MCi_STATUS value.
Arg4: 0000000001010135, Low order 32-bits of the MCi_STATUS value.

9: kd> !errrec ffff8b8cdbe7b028
===============================================================================
Common Platform Error Record @ ffff8b8cdbe7b028
-------------------------------------------------------------------------------
Record Id : 01d9670ab6ceff53
Severity : Fatal (1)
Length : 936
Creator : Microsoft
Notify Type : Machine Check Exception
Timestamp : 4/4/2023 16:17:46 (UTC)
Flags : 0x00000000

Section 0 : Processor Generic
-------------------------------------------------------------------------------
Descriptor @ ffff8b8cdbe7b0a8
Section @ ffff8b8cdbe7b180
Offset : 344
Length : 192
Flags : 0x00000001 Primary
Severity : Fatal

Section 1 : x86/x64 Processor Specific
-------------------------------------------------------------------------------
Descriptor @ ffff8b8cdbe7b0f0
Section @ ffff8b8cdbe7b240
Offset : 536
Length : 128
Flags : 0x00000000
Severity : Fatal

Minidump File Download Link (if you want analysis too) :
https://drive.google.com/file/d/1XHzpIHFWndqFmB0VUM2-jrhb5xU6l40w/view?usp=sharing

After them i tested my CPU with Prime95 Small Test FFT. I closed PBO and CBS from bios. I didn't get any BSOD. Here is test results.

PS : I updated my bios to latest version after this test.

Lastly,

I'm using the CPU with CBS, PBO turned off. XMP is turned off for ram. I deleted the drivers with DDU and installed the latest version from scratch. I scanned the system with scf /scannow via cmd and provided the fixes. I played GTA V and RDR2 games for 30 minutes each and it didn't give a shutdown or blue screen condition.

But there is a situation like this, in 2 games, while VSYNC is turned off, I experience FPS Drop in some cases. This goes away when I turn on VSYNC. As far as I guessed, it gets into a bottleneck because the CPU's Voltage and Ghz are low. As you can see in the pictures below, the Processor works between 1.0V and 2.5-3.0 Ghz. The clock of the video card is also in the 1980 - 1990clk.

First image is GTA V, second image is RDR2.

Could it be a hardware problem with the video card? Could it be on the processor? XMP is turned off, is the ram putting a power on the system right now? Or is it a software problem?
Waiting for your valuable feedback on the subject.

Regards.

 

[3x0]

It's possible that the CPU is failing under light load scenarios. You've only tested heavy load. Download https://github.com/sp00n/corecycler/releases/tag/v0.9.3.0 and let it run at least for 10 cycles, close all background apps while testing.
Do you have Resizable BAR enabled?

 

[Berfs1]

If you tested the RAM with XMP on, leave it on, it's not the RAM causing problems here. If you have any tuning options set for the CPU, revert them and run the CPU completely stock, and consider completely reinstalling windows. It's possible your OS is corrupted beyond repair, as was the case with my 11900K build recently. It happened because I was tuning ram within windows, and with the unstable overclocks it was rapidly corrupting the OS, and I have since moved to testing in memtest86 as it is built in to my BIOS and not in the OS. Because of the massive corruption, mobile hotspot wasn't even working properly, my VMs started to fail, and lastly I couldn't even get into windows, that's how bad it got, and I suspect it could be a similar case for you too.

Edit: About the RAM, I see in screenshots you are running 3000 CL16 20-20-38 1T, but your specs are listed as 3200 8x8GB?
Edit 2: In regards to what 3x0 is mentioning about crashing under light load scenarios, that was actually also happening to me on the 11900K whenever I had my (what I thought was) stable overclocks running, it was randomly restarting even when idle, so I reverted it to stock and just adjusted power limits. If your CPU is failing even under stock and even with a new BIOS and new OS, I would consider doing an RMA.

 

[000me]

It's possible that the CPU is failing under light load scenarios. You've only tested heavy load. Download https://github.com/sp00n/corecycler/releases/tag/v0.9.3.0 and let it run at least for 10 cycles, close all background apps while testing.
Do you have Resizable BAR enabled?

No resizable bar not enabled, actually i didn't do any changes for it. So i don't know enabled or disabled. I'm so sorry but i'm not trust this no name scripts. Can you offer me another program to test it please. @3x0

If you tested the RAM with XMP on, leave it on, it's not the RAM causing problems here. If you have any tuning options set for the CPU, revert them and run the CPU completely stock, and consider completely reinstalling windows. It's possible your OS is corrupted beyond repair, as was the case with my 11900K build recently. It happened because I was tuning ram within windows, and with the unstable overclocks it was rapidly corrupting the OS, and I have since moved to testing in memtest86 as it is built in to my BIOS and not in the OS. Because of the massive corruption, mobile hotspot wasn't even working properly, my VMs started to fail, and lastly I couldn't even get into windows, that's how bad it got, and I suspect it could be a similar case for you too.

Edit: About the RAM, I see in screenshots you are running 3000 CL16 20-20-38 1T, but your specs are listed as 3200 8x8GB?
Edit 2: In regards to what 3x0 is mentioning about crashing under light load scenarios, that was actually also happening to me on the 11900K whenever I had my (what I thought was) stable overclocks running, it was randomly restarting even when idle, so I reverted it to stock and just adjusted power limits. If your CPU is failing even under stock and even with a new BIOS and new OS, I would consider doing an RMA.

Yes i tested with XMP on so i'm turning on it. I didn't understand this ;

It's possible your OS is corrupted beyond repair, as was the case with my 11900K build recently. It happened because I was tuning ram within windows, and with the unstable overclocks it was rapidly corrupting the OS, and I have since moved to testing in memtest86 as it is built in to my BIOS and not in the OS.

I didn't any overclock on my cpu. Actually i didn't do anything on BIOS i just changed XMP to "profile 1" 3000mhz. I upraged my system 2 months ago and up to now i didn't play any games. But i wanted to play game and get this bsod. After i closed my PBO and CBS i lost these bsod errors. But now i'm getting fps drops. I'm not getting any random bsod errors. So you suggest me to reset all my bios settings? Let me test it like that? @Berfs1

PS: Sorry i wrote it wrong. Yes RAM's 3000mhz CL16 8x8.

 

[droopyRO]

I guess you don't have a spare AM4 CPU to test with ? I had a 5600X back in 2020 when it launched. It gave me WHEA errors when getting in or comming out of Sleep Mode. I even had a thread here iirc. Ended up swapping the mobo, GPU, RAM kits, PSU, reinstalling Windows. Then i RMA the 5600X, after about a week i got my money back. Don't know what your issue is, but checking it out with another CPU, PSU, RAM kit and even GPU would be the way to isolate it. BTW do you have the latest BIOS for your motherboard and/or did you try to reset your BIOS with you PC unplugged ? Also that 650W PSU is what brand exactly ?

 

[000me]

I guess you don't have a spare AM4 CPU to test with ? I had a 5600X back in 2020 when it launched. It gave me WHEA errors when getting in or comming out of Sleep Mode. I even had a thread here iirc. Ended up swapping the mobo, GPU, RAM kits, PSU, reinstalling Windows. Then i RMA the 5600X, after about a week i got my money back. Don't know what your issue is, but checking it out with another CPU, PSU, RAM kit and even GPU would be the way to isolate it. BTW do you have the latest BIOS for your motherboard and/or did you try to reset your BIOS with you PC unplugged ? Also that 650W PSU is what brand exactly ?

Yes i have latest BIOS but no didn't to try reset with PC unplugged. PSU is POWER BOOST VK-G3403S 650W 80+. No i don't have spare CPU, RAM, GPU or PSU. You suggest me RMA to CPU? CPU can be made with problems?

Thanks for all your helps.

PS: I will try your steps on previous answer.

 

[droopyRO]

PSU is POWER BOOST VK-G3403S 650W 80+

That seems to be a case + PSU combo, not the PSU model. While it might work with that no-name PSU, that system is pulling around 300-400W from the wall, depending on the efficacy of he PSU itself.
I can't suggest you do anything, since there is no way to isolate your problem :/

 

[RJARRRPCGP]

If you tested the RAM with XMP on, leave it on, it's not the RAM causing problems here. If you have any tuning options set for the CPU, revert them and run the CPU completely stock, and consider completely reinstalling windows. It's possible your OS is corrupted beyond repair, as was the case with my 11900K build recently.

Looks like a faulty CPU, faulty motherboard or a faulty PSU. Based on error information given.

 

[000me]

Looks like a faulty CPU, faulty motherboard or a faulty PSU. Based on error information given.

That seems to be a case + PSU combo, not the PSU model. While it might work with that no-name PSU, that system is pulling around 300-400W from the wall, depending on the efficacy of he PSU itself.
I can't suggest you do anything, since there is no way to isolate your problem :/

Hello,

As you said, I ran Cinebench and Furmark, Prime95 tests again with XMP turned off but CBS and PBO settings turned on. There was no problem during the tests, I did not get any errors, as in the previous one.

But when I tested it in RDR2 and GTA V games, I got Whea_Uncorrectable error again after 10 minutes in RDR2. But there was no FPS Drop as far as I could see.

When I tried GTA V in the game, FPS Drop was entering unbelievably. When I turn on VSYNC, there is no FPS Drop. In order to push it a little more later, I pulled the settings such as MSAA, FXAA, Shadow etc. to the last with VSYNC turned off. At first there seemed to be no problem, but of course the temperature values started to increase, then it stopped at those temperatures and did not increase. After continuing the game like this for 10 minutes, the game and the computer got locked and restarted itself.

In below I have listed the test images one by one by name. Along with GTA V and RDR2 screenshots, I prepared a video showing the GTA V FPS Drop and VSYNC event at the bottom of the subject. (I edited the video and uploaded it to Youtube so that it wouldn't take too much of your time.)

Thank you for all your help.
@droopyRO @RJARRRPCGP @Berfs1 @3x0

PS : PSU Model is Power Boost BST-ATX650A

Cinebench Mutli Core Test

Cinebench Singe Core Test

Prime95 Test

Furmark Test

RDR2

GTA V

GTA V Menu is laggy. (Look the GPU and CPU Usage)

GTA V Test Video.

 

[RJARRRPCGP]

Your core temps are absolutely terrible for a 5600/X! When running Prime95 and the non-game benches. Actually surprised the CPU usage isn't at least around 100 percent in the GTA V menu one.

 

[Kodehawa]

How do you get 81C while the CPU draws 30W? Bad CPU mount or horrible cooler. CPU is likely not working properly because some area is too hot.

 

[emothxughts]

I bought a Sapphire Pulse RX 6700XT graphics card.

Is this bought new or used? I had a WHEA 18 Cache Hierarchy Error once, I thought it was my R5 3600 CPU that failed, so I RMA'd it, I still get the same error with my replacement CPU. After lots of stress testing, I found out the culprit was the used Gigabyte RX5600XT card I bought. (I have a thread detailing this situation, search my profile)

I know you mentioned not having a spare GPU, but can you take the PC to a computer shop and have them test it out with their GPU?

 

[3x0]

When I tried GTA V in the game, FPS Drop was entering unbelievably. When I turn on VSYNC, there is no FPS Drop.

GTA5 engine will break/stutter when the FPS exceeds ~150FPS or something similar. It's not abnormal

 

[RJARRRPCGP]

Hello,

As you said, I ran Cinebench and Furmark, Prime95 tests again with XMP turned off but CBS and PBO settings turned on. There was no problem during the tests, I did not get any errors, as in the previous one.

But when I tested it in RDR2 and GTA V games, I got Whea_Uncorrectable error again after 10 minutes in RDR2. But there was no FPS Drop as far as I could see.

When I tried GTA V in the game, FPS Drop was entering unbelievably. When I turn on VSYNC, there is no FPS Drop. In order to push it a little more later, I pulled the settings such as MSAA, FXAA, Shadow etc. to the last with VSYNC turned off. At first there seemed to be no problem, but of course the temperature values started to increase, then it stopped at those temperatures and did not increase. After continuing the game like this for 10 minutes, the game and the computer got locked and restarted itself.

In below I have listed the test images one by one by name. Along with GTA V and RDR2 screenshots, I prepared a video showing the GTA V FPS Drop and VSYNC event at the bottom of the subject. (I edited the video and uploaded it to Youtube so that it wouldn't take too much of your time.)

Thank you for all your help.
@droopyRO @RJARRRPCGP @Berfs1 @3x0

PS : PSU Model is Power Boost BST-ATX650A

Cinebench Mutli Core Test
View attachment 291170

Cinebench Singe Core Test
View attachment 291171

Prime95 Test
View attachment 291172

Furmark Test
View attachment 291173

RDR2
View attachment 291174

GTA V
View attachment 291175

GTA V Menu is laggy. (Look the GPU and CPU Usage)
View attachment 291176

GTA V Test Video.

That video, for the first vehicle, reminds me of when there's packet loss with online gaming.

 

[3x0]

That video, for the first vehicle, reminds me of when there's packet loss with online gaming.

It's not, just the GTA5 engine being buggy at high framerates.

 

[RJARRRPCGP]

It's not, just the GTA5 engine being buggy at high framerates.

Even for 144 FPS? Yikes! I wonder if mine will remain smooth, OTOH. I guess I should test.

 

[droopyRO]

Thank you for all your help.

Did you try to run the PSU test from OCCT or run Furmark and Prime95 at the same time ? If your PC freezes or reboots when both CPU and GPU are at 100% then your PSU might be your issue. Do you have a warranty seal on your case ? if not, try to take a picture of the PSU sticker and the inside of your case.

 

[000me]

Is this bought new or used? I had a WHEA 18 Cache Hierarchy Error once, I thought it was my R5 3600 CPU that failed, so I RMA'd it, I still get the same error with my replacement CPU. After lots of stress testing, I found out the culprit was the used Gigabyte RX5600XT card I bought. (I have a thread detailing this situation, search my profile)

I know you mentioned not having a spare GPU, but can you take the PC to a computer shop and have them test it out with their GPU?

It's new. Well i'm still thinking it failed because of CPU. Minidump file says that no i'm. You got different error. But after i take new parts i will check it too.
Thanks.

Your core temps are absolutely terrible for a 5600/X! When running Prime95 and the non-game benches. Actually surprised the CPU usage isn't at least around 100 percent in the GTA V menu one.

How do you get 81C while the CPU draws 30W? Bad CPU mount or horrible cooler. CPU is likely not working properly because some area is too hot.

Did you try to run the PSU test from OCCT or run Furmark and Prime95 at the same time ? If your PC freezes or reboots when both CPU and GPU are at 100% then your PSU might be your issue. Do you have a warranty seal on your case ? if not, try to take a picture of the PSU sticker and the inside of your case.

Hello,

Today i order new PSU, CPU Cooler and new fans. No i din't run PSU test because i'm little afraid with this PSU. It is not much good.

When my orders come, I will assemble them. Then I do the PSU test and other tests again. I'll post the results here when the tests are done. I hope the problem is just temperature and power supply related.

This is my order list

Corsair RM750 - 750W 80PLUS GOLD ( CP-9020231-EU ) - Power Supply
ID-COOLING SE-224-XT White V2 - CPU Cooler
ID-COOLING XF-12025-SD-K - 120mm Fan

PS : I really spent a lot of time researching these. What do you think about this parts too?

Thanks for everything again. @RJARRRPCGP @droopyRO @3x0 @Kodehawa @emothxughts

 

[000me]

Did you try to run the PSU test from OCCT or run Furmark and Prime95 at the same time ? If your PC freezes or reboots when both CPU and GPU are at 100% then your PSU might be your issue. Do you have a warranty seal on your case ? if not, try to take a picture of the PSU sticker and the inside of your case.

Even for 144 FPS? Yikes! I wonder if mine will remain smooth, OTOH. I guess I should test.

It's not, just the GTA5 engine being buggy at high framerates.

How do you get 81C while the CPU draws 30W? Bad CPU mount or horrible cooler. CPU is likely not working properly because some area is too hot.

Hello again,

I received the parts I ordered and I assembled them all. First of all, if there are people who are thinking of buying a PSU or CPU cooler, they can buy these products, they are very satisfied with the products. I chose Arctic MX-4 as thermal paste and decided that it was the best in terms of price and performance. I searched 170 forums / topics in order not to have problems with the products.

A Perfect Guide for Those Who Want to Buy a PSU: https://cultists.network/140/psu-tier-list/

I did Prime95, Furmark, Cinebench tests and there was no problem again. I experienced an incredible drop in CPU temperature. As you can see from the screenshots below, it does not exceed 65 degrees at full power. In games (with Ultra graphics settings) the CPU is between 50-55 degrees.

As you remember from the topic title, I was getting WHEA_Unccortable error in the RDR2 game. After replacing these related parts, I played for 30-35 minutes, but I did not experience any problems (BSOD) or FPS Drops. Maybe I got the solution for the issue. I think it is due to the unstable operation of the processor as it cannot provide enough power to the processor at high power consumptions due to the fact that the event is completely overheating and I use a poor quality power supply.

For GTA V, I still have FPS Drop problem. When I activate VSYNC, this problem disappears. As far as I have observed, it is due to the fact that the FPS suddenly drops to 120 when I go to the place where there are indoor or few objects and go to the FPS 180, and then suddenly enter the city or traffic. Actually, 120-140 FPS is fine for Ultra Graphics Settings. I still don't understand this issue...

I share screenshots of tests and games below. I am sharing a 1-minute video I edited about GTA V FPS Drop.

Thanks for all your helps. @RJARRRPCGP @3x0 @Kodehawa @droopyRO @emothxughts

Prime95 30 Minute Test


CineBench Multi-core Test


Furmark 15 Minute Test


Red Dead Redemption 2 (SS from played 30 minute)


GTA V (SS from played 30 minute)


GTA V New FPS Drop Video (19/04/2023) :

 

[Kodehawa]

For GTA V, I still have FPS Drop problem. When I activate VSYNC, this problem disappears. As far as I have observed, it is due to the fact that the FPS suddenly drops to 120 when I go to the place where there are indoor or few objects and go to the FPS 180, and then suddenly enter the city or traffic. Actually, 120-140 FPS is fine for Ultra Graphics Settings. I still don't understand this issue...

That's the engine's fault, nothing to do with your system. The game does not like to go over 180fps and will stutter really hard once it does.

 

[Berfs1]

If adding a more powerful cooler fixed your WHEA errors, it is fixing the problem for now, but keep in mind it will come up again soon (like within a year), because the CPU is likely degraded to the point it can't handle its stock spec even at 80C anymore. All CPUs degrade over time, so eventually it will not be able to run stock spec after a certain point, so if it is fixed with a new cooler but still crashes with the old cooler, I highly recommend you still do an RMA while you can, or a refund and get yourself another chip, because it will cause issues sooner than you will expect.

 

[000me]

That's the engine's fault, nothing to do with your system. The game does not like to go over 180fps and will stutter really hard once it does.

Then i'm not care anymore GTA's FPS Drop. I can't do anything about that. @Kodehawa

If adding a more powerful cooler fixed your WHEA errors, it is fixing the problem for now, but keep in mind it will come up again soon (like within a year), because the CPU is likely degraded to the point it can't handle its stock spec even at 80C anymore. All CPUs degrade over time, so eventually it will not be able to run stock spec after a certain point, so if it is fixed with a new cooler but still crashes with the old cooler, I highly recommend you still do an RMA while you can, or a refund and get yourself another chip, because it will cause issues sooner than you will expect.

Your message has me very worried. I want to RMA the product, but I did not do anything for now because the process is too long. @Berfs1

I wonder what other people think. Thanks.

 

[RJARRRPCGP]

The game does not like to go over 180fps and will stutter really hard once it does.

I saw it randomly pausing at less than 180 FPS. I dunno. Even at around only 144, IIRC!

 

[000me]

I saw it randomly pausing at less than 180 FPS. I dunno. Even at around only 144, IIRC!

Should i RMA it? What do you think too. @RJARRRPCGP

 

[RJARRRPCGP]

Should i RMA it? What do you think too. @RJARRRPCGP

More likely, you should if it's a 6700 XT. Or return it of course, if you suspect being scammed by a miner. (More likely with a pre-6750!) (anything before 6750 XT)