Question Recurring BSODs after a month of troubleshooting and full hardware/software replacement ?

[Fadi.Alhassi]

Hello!
This is my first time ever trying online forums, I am actually doing this because I am desperate and drained after a full month of hardcore troubleshooting trying to fix things on my own
I will try to be concise as much as possible, because it is a long story unfortunately.

I built a new gaming PC 1 month ago, and I have been having recurring BSODs ever since.

Initial build:
CPU: 7600x
GPU: 7900 xt Saphire
Ram: 16 GB (8x2) Kingstone fury beast 5200Mhz
PSU: Cooler Master MWE 850 Gold V2 Fully Modular PSU (UK Plug) - 80 PLUS Gold 850W Power Supply Unit, Flat Black Cabling, 120mm HDB Fan, High-Temperature T
HDD: Crucial P3 Plus 1TB M.2 PCIe Gen4 NVMe Internal SSD - Up to 5000MB/s - CT1000P3PSSD8
MOBO: MSI B650 Pro Wifi

I had recurring BSODs, tried both windows 10 and 11, both were fresh installs.
Reinstalled all drivers
Tried all BIOS versions
Tried XMP on and OFF
Reset BIOS settings to factory default, no luck with any of that, the frequency and type of BSODs might change overtime, specially after BIOS updates, but it never stopped

After all that Software troubleshooting, and some online searching, I thought it could be a hardware proplem.

First:
I replaced both my RAM and my PSU with the a similar brand new parts and reinstalled windows with complete data wipe; no luck

Then I replaced my MOBO to B650 Tomahawk thinking my MOBO might have been the proplem, with a fresh install of Windows and drivers, no luck.

Then: I switched my GPU to 4070 ti thinking it was an AMD driver issue, again with fresh windows install, no luck, so reverted back to 7900 XT

then:
I replaced my HDD to Kingstone SSD with fresh windows install and both disabling and enabling XMP, again no luck
Reverted back to Crucial

And finally, I replaced my Ram for the second time, this time switching to Crosair Vegnacne 32 GB (16x2) 6000Mhz AMD Expo compatible (thinking that might be the issue) and tried it with both Exp and stock settings. Again, no luck.

I tried couple of BIOS versions on this final setup, problem persisted
Currently on the latest version 1.31 (9-Jan)

As I formmatted my PC around 20 times, I have only my most recent dump files, so I'll attach them here.

Please note that I replaced every single piece of hardware in my PC except for the CPU and the CPU fan
And I also tried every software, online troubleshooting, and google searching there is (that I am capable of and within the domain of my humble IT knowledge) with no luck.

Anyone have any idea what the problem might be?

As much as I would love my new and expensive PC to work, but TBH, at this point, I am also curious to know what kind of problem is that? I mean how come it is still happening after trying every Software;Hardware combination possible, and almost replacing my hardware entirely with no luck!
All I can think of is either hardware incompatibility,
Or possibly a faulty BIOS driver, as I always noticed a change as I tampered with that. But still many people are using a smiliar build to mine, I can't imagine they're all having the same problem because my compute is unusable!

Any thoughts?


Minidump files:
https://drive.google.com/file/d/1_TcieLvLZUqjX5OlsSZOX_ry1ffQ4-Bx/view?usp=sharing

 

[johnbl]

my first guess would be your graphic card or driver is the cause of the problem. Remove any software overclock or maybe even underclock the GPU to the stock rate. I would confirm that each connection to the gpu is proper. IE pins not pushed out of supplemental power connectors. Confirm that the current drivers are installed.

you should also find out what driver owns the pooltag
202b
use this method:JSI Tip 7102. How do I determine a driver name from a pool tag? | ITPro Today: IT News, How-Tos, Trends, Case Studies, Career Tips, More
something like this command
findstr.exe /m /l "202b" *.sys

if it finds the string 202b in that directory then it will return the file name it was found in. I expect it will be a amd graphics driver and I also think they store them in their own directory so the command may not find any driver unless you enter into the amd gpu driver directory (not sure of its location)

here is a known list of pooltags:
PoolMonX/pooltag.txt at master · zodiacon/PoolMonX · GitHub (i did not see any tags that start with a number)

I would also update bios or go and reset it to default and make sure they are not overclocking the PCI bus
(not over 100MHz) save the updated bios and reboot.



-------------------------------
bugchecks most current first:
will have to look at the other bugchecks
==========
4th bugcheck
basically a stack overflow in the direcx driver.
maybe a double remove of an object.


===========
3rd bugcheck
system bugcheck running steam, got an access violation in the windows kernel.

============

second bugcheck:
access violation in windows display manager
function that failed was dereferencing some directx function.
system was running for 15 minutes
unknown pool tag = 202b

(pooltags are required to identify the driver)


------------
1st bugcheck was a access violation in TM.sys
Kernel Transaction Manager Driver file
system was running for only 39 seconds and was trying to read the registry at the time of the access violation that caused the bugcheck.

-none of the windows files are corrupted.

BIOS Version 1.10
BIOS Release Date 10/13/2022
Manufacturer Micro-Star International Co., Ltd.
Product MAG B650 TOMAHAWK WIFI (MS-7D75)

Processor Version AMD Ryzen 5 7600X 6-Core Processor
Processor Voltage 8dh - 1.3V
External Clock 100MHz
Max Speed 5450MHz
Current Speed 4700MHz

 

[Fadi.Alhassi]

I really appreciate that you took the time; thank you so much!
But before doing any of that, it is worth mentioning that at some point in my troubleshooting, I got a brand new 4070 Ti, installed windows with full data wipe and installed Nividia drivers, and the problem wasn't solved. So do you still think it is a GPU driver issue keeping that in mind?

I also tried underclocking the GPU before, no luck
And I'm running everything on stock settings, no overclocking what so ever

 

[johnbl]

I really appreciate that you took the time; thank you so much!
But before doing any of that, it is worth mentioning that at some point in my troubleshooting, I got a brand new 4070 Ti, installed windows with full data wipe and installed Nividia drivers, and the problem wasn't solved. So do you still think it is a GPU driver issue keeping that in mind?

I also tried underclocking the GPU before, no luck
And I'm running everything on stock settings, no overclocking what so ever

I would expect it to be the integrated GPU inside the CPU.
Can it be disabled? and only the dedicated gpu enabled?
be sure to go into bios and toggle any setting and save config after new hardware is added. This will force the bios to rebuild the database it sends to windows when the machine boots up. IE it tells windows what resources are in use by the bios.
guessing the cpu gpu display driver is
AtihdWT6.sys Tue Sep 13 11:22:21 2022

nvidia driver was not installed in the most current dump. (if the nvidia hardware is in the machine the bios will have reserved the primary gpu resources)

 

[Fadi.Alhassi]

I would expect it to be the intergraded GPU inside the CPU.
Can it be disabled? and only the external gpu enabled?
be sure to go into bios and toggle any setting and save config after new hardware is added. This will force the bios to rebuild the database it sends to windows when the machine boots up. IE it tells windows what resources are in use by the bios.
guessing the cpu gpu display driver is
AtihdWT6.sys Tue Sep 13 11:22:21 2022

nvidia driver was not installed in the most current dump. (if the nvidia hardware is in the machine the bios will have reserved the primary gpu resources)

I thought of that at some point as well, I tried disabling it from both BIOS and windows (2 different attempts); no luck

Yeah Nividua driver is not there because I returned the 4070 Ti since I came to the conclusion it's not an AMD issue

 

[Fadi.Alhassi]

Here is fresh Minidump file I just had with interated graphics disabled through BIOS (also double checked it is not being detected by windows):
https://drive.google.com/file/d/1ZLFdzpSWPK9Df4iGqAibw0yVQpciS7sy/view?usp=sharing


Could this be the CPU? Because it is literally the only hardware part I didn't replace..

 

[johnbl]

edit: looks like this motherboard has a mediatek wireless 6e wifi and bluetooth. I could not find the actual chip number but I do see that some vendors are pushing out a fixed driver dated 1-18-2023. At least that was the file date, I am not sure of the driver date.
some vendors may take some time to put the update on their website. I guess as a side note, if you have a 6e router you should make sure its firmware is current.
bugs in a router can corrupt packets and cause retransmit request that cause 300 second network backups in processing packets.

this bugcheck was due to a DPC taking too long.
with a minidump I can only look at the DPC queue on the active processor, in this case core 3. I see network calls and calls to
stornvme. I can not see the other 11 cores.
to see them you need to change the memory dump type to kernel then provide the memory.dmp file.

system was up for 7 minutes, game running.
best guess based on the one core would be to update firmware in your ssd and update your network adapter or turn off any that are not being used. you can download crystaldiskinfo.exe to read the firmware version of your drive then look to see if there is a update.
(none of the dpc were active)

network driver:
realtek 2.5 gb pci (wired driver)
rt25cx21x64.sys Fri Mar 18 02:24:07 2022
NetAdapterCx.sys (microsoft network class extension driver)

also don't forget to get the amd bluetooth driver update
if there is one. (part of the network drivers)


MediaTek WiFi device driver
mtkwl6ex.sys Fri Jan 7 01:39:55 2022
mtkbtfilterx.sys Fri Jan 7 00:54:24 2022
(my guess would be this device is causing the delay, but i have no proof)
you might consider disabling and removing the media tek wifi driver and using the wired network and see if you have the same problem.
(since you crash so fast anyway)

otherwise change to the kernel dump and I can see the other cores and internal logs.

no other error indications, just a backlog on processing driver request

edit:
MAG B650 TOMAHAWK WIFI (msi.com)
shows a updated driver for wireless network dated
10-3-2022

and 12-30-2022 for the wired driver.

you should update both and retest.
new chipset drivers came out on 2023-02-13
for your motherboard

note: running
AMDRyzenMasterDriver.sys Thu May 12 23:01:00 2022

expect new updates for the new rounds of security fixes. You have to manually install when they become available.

 

[Fadi.Alhassi]

I am already using Ethernet cable, and tried Wifi before as a part of my troubleshooting.

Final update:
I replaced my CPU today with another 7600x as a last hope, given that it is the only hardware part I didn't replace even though no crash report was signaling a faulty CPU and it was doing fine on all CPU benchmarks with no crashes, and guess what, it seems to be working!!

I have been gaming for 3 hours with no crashes!

Thank you so much johnbl for your time and effort.
Will keep you updated if anything new comes up.