Consistent WHEA BSoD's

Question

Hello!

Finally decided to take the dive and share my frustration over a consistent but weird series of blue screens,

Specs:

Mobo:

MSI MS-7C37 (Updated Bios)

CPU:

Ryzen 5800x (Cooled by an AOI)

Memory:

32 GB - DDR4 memory - Corsair (4x8gb)
GPU:

Gigabyte RTX 3090 Turbo (No longer supported by gigabyte + stock bios)

OS:

Win 10 Pro Version: 22H2 (windows updates are a hit-miss when it comes to stability)

Drives:

2x NVME SSD's - games
2x 2TB HDD's - storage / programs
1x SSD (sata) - os installed on here

PSU:

MSI MPG 1000 watt

Problem:

Simplest way to put it is with the RTX 3090 i've had nothing but semi-consistent BSOD's, when using the GTX 1060 on the other hand there's none.

I'm at a loss for what even causes it after trying many proposed solutions.

What happens before a BSOD over the various instances: (in order of how common they are)

• Screen freezes for 1-2 seconds - blue screen - fairly slow data collection - reboots (most recently and commonly)
• Straight up BSOD - quick data collection - reboots (fairly common)
• Full black screen - reboots instantly (Every few bsod's this happens) - Produces BSOD logs however

Tried solutions:

Mobo Bios updated

Mem check - nothing found

Storage check - nothing found
GPU / OS updates -every few drivers delays it for up to 14 days or increases it to every hour - Every GPU driver gets installed "cleanly" through nvidia's software

Underclocking the gpu - no results

Increasing the Temp limit from 83C to 91C on the GPU - thought maybe it overheats and shuts down when it hits 83. (it's back at 83 after a failed attempt)

Changed the GPU FAN's speed relative to it's Temperature (it hits max RPM sooner at around 75C when before it was set at 80C)

Tried looking for a GPU Bios update but it's unsupported by Gigabyte and their support team is awfull - there exist driver updates on third party websites but i can't exactly afford to potentially BRICK this gpu.

PSU replaced, originally a 850 watt and this is now swapped for a 1000 watt - while the swap happened i've also checked all connections to make sure they're secure, including fans, sata, power, etc..

Some observations around the crashes:

Unsure about the CPU as at most 4-6 cores are being used between 20-40% around the times these BSOD's happen.

Memory usage hovers around 12/32 GB

Disk usage is minimum

GPU usage is usually at 100% (with a powerdraw of around 350 Watt)

Games/Programs it tends to crash with most recently:

Netflix app / Netflix on chrome (surprisingly)

WH40k Darktide (optimized settings)

Icarus (maxed settings)

Deep Rock Galactic (max settings)

Cyberpunk (raytracing enabled)

Final Fantasy XIV Online (max settings)

Programs usually open during crashes:

Chrome

Discord

Steam

Dump files (4 days, 6 files Zipped up): https://drive.google.com/file/d/1KHpl8qRBm-VTVs_nvKfzsqhr-Q3KyE7o/view?usp=share_link

Additionally:

I'm fully aware that a full reinstall of windows could be a fix, but i don't want to resort to this if the issue might be something simple as some kind of driver incompatibility which i really hope it is.

I also can't afford it to replace the more expensive parts easily unlike something smaller such as the psu, memory or drives.

Risking damaging the CPU especially with AMD by checking the pins is also something i don't fancy doing.

All of these would literally be a last ditch effort for a fix.

Thank you in advance whoever manages to find a solution.

Answer 1

Hi V.Jansenm, okay, i will be here.

Answer 2

Hello!

I will be trying the proposed solutions, note that it might take a little while considering the inconsistent rate of them.

Just in case i'll backup any further bsod's with the profile changes then after that i'll run the verifier as well.

Answer 3

Hi V.Jansenm

Please try his process, if this does not find the underlying cause, then I believe you may have a hardware problem.

To try to force Windows 10 show any faulting drivers or other cause, the best option would be to turn on Driver Verifier, let your PC crash 3 times, then you must turn off Driver Verifier, and finally, upload any newly created minidump files

Before you run Driver Verifier, please create a new System Restore Point

https://answers.microsoft.com/en-us/windows/for...

---

Note, if you have any difficulty getting into Windows with Driver Verifier enabled:

Start your PC, just as Windows attempts to load (spinning dots), press and hold Power Button for 5 - 10 seconds to perform a Hard Shut Down

Do this twice

On the third start Windows will boot into the Recovery Environment and from there you can access System Repair, Safe Mode, Command Prompt... etc.

Go to Troubleshoot - Advanced Option - Startup Settings and click Restart

Upon restart, press 4 to enter Safe Mode

Open Command Prompt as Administrator and run these two commands, then restart your PC.

verifier /reset

verifier /bootmode resetonbootfail

Answer 4

Hey,

Thanks for the swift reply first of, very much appreciated.

Interestingly enough, i didn't expect that one actually, the cpu is left as is at stock speeds however.

Memory used to run at 2133 before but is now running at it's rated speed of 3600 after the bios update, i believe those profiles are disabled i checked them earlier today.

I suppose i forgot to clarify that one detail in everything my apologies.

Answer 5

Hi V.Jansenm

I am Dave, I will help you with this.

Your GPU is not listed as a contributory factor to the system crashes, the minidump files all indicate it is a driver associated to your AMD processor that is causing the system crashes.

Are you overclocking or undervolting your CPU?

Is there an XMP/DOCP profile set on your RAM in BIOS, o are you running your RAM at its default clock speed?