BSOD Nightmare, new build

[thekingswolf]

Build:

MB: ROG STRIX Z690-E
BIOS: 1403
Firmware: V007

CPU: Intel Core i9-12900K (BX8071512900K)
RAM: G.SKILL Trident Z5 RGB Series 32GB (2 x 16GB) DDR5 6000 Intel XMP 3.0 (F5-6000J3636F16GX2-TZ5RK)
GPU: GIGABYTE Gaming GeForce RTX 3070 Ti 8GB GDDR6X PCI Express 4.0 (GV-N307TGAMING OC-8GD)
HD: SAMSUNG 980 PRO PCIe® 4.0 NVMe m.2 (MZ-V8P1T0B)
PSU: Seasonic FOCUS Plus 1000 Gold SSR-1000FX 1000W 80+ Gold 
FrontPanel Addon: EZDIY-FAB 5.25in 10Gbps USB 3.1 Gen2 Hub and Type-C Port,Front Panel USB Hub with QC 3.0 Quick Charging, connected to MB via U32G2X2_3

GPU installed in first PCI slot #1, no riser
M.2 installed in slot 1
Addidional drives connected via SATA6G_1 and SATA6G_2

*Please note that during all stages of troubleshooting, Windows Updates have been run until no further updates available.

Build complete, Windows 11 Pro format. Had no problems for a few hours, but after my set up was complete and I was gaming on it, I started geting blue screens. BSOD occured randomly, gaming, not gaming, watching youtube, listening to music, it was not predictable. Each BSOD was a diferent message, but after review of the minidumps there were 2 causes. One was apparently the onboard ethernet (SYSTEM_THREAD_EXCEPTION_NOT_HANDLED ndis.sys+796ac), which I disabled and installed my Asus XG-C100C 10gb ethernet card in bottom PCI slot #4. I have not gotten another BSOD with that as the cause since disabling the onboard ethernet. The next cause of the blue screen is the one that has been persistent, and is listed here along with the different BSOD messages I have seen:

ntoskrnl.exe ntoskrnl.exe+4295a9 fffff800`16a00000 fffff800`17a47000 0x01047000 0x080ec80a (WinBuild.160101.0800) Microsoft Corporation C:\WINDOWS\system32\ntoskrnl.exe

BSOD's so far
system_service_exception
irq_not_less_or_equal - happens instantly pre drivers, and randomly post drivers
page_fault_in_non_paged_area
DRIVER_OVERRAN_STACK_BUFFER
UNEXPECTED_KERNEL_MODE_TRAP


Knowing drivers are usually the problem with these things, thats what I messed with first. My first install, the Armory Crate was auto installed, and I used the drivers from that. So next I manually downloaded all the drivers direct from the ASUS support page for the board and installed those instead, no change. Blue screens continued.

So I did a clean format and started from scratch, this time I went back to Windows 10 Pro and only used the drivers from the support page instead of Armory Crate. Same concerns, had to disable onboard ethernet again, ntoskrnl.exe BSOD persists. I ran a 6hr full MemTest86 and the RAM tested all good. Firmware for the Samsung m.2 drive was available so that was done as well.

Beacuse of that one page fault error, and having eliminated the RAM as the problem via memtest86, I wanted to eliminate the m.2 drive or the slot as the possible problem. I moved the m.2 drive to a different slot and started over, same results. I used a different m.2 drive I have as a space in a little external enclosure I use sometimes. Different m.2 drive has the same issues in all m.2 slots. In my mind, this eliminates the m.2 drives themselves, but not necessarily the board slots/drivers.

Wanting to get away from the m.2 slots, I removed it the m.2 drive completely and connected an SSD to SATA6G_1 for the OS, and moved my two other drives to SATA6G_2 and SATA6G_3. Clean format to Windows 11 Pro again since Windows 10 was having same problems. The first thing I do after a windows install is change the locations of all my personal folders. Desktop, Documents, Pictures, Music, and Videos are all moved to D:\ folders. For this install I decided not to move those folders in case there was some problem with those drives or the SATA ports, so I left them at thier default locations. Another note, I have an 8tb drive as E:\ that I use for game installs and my downloads folder. When reinstalling Steam to the same folder I always use on this drive, I have noticed that the install is lightning fast like all the others, but for some reason the final step where the install is just creating a shortcut to the executable takes a long time. Like, several minutes. The entire install takes maybe 15 seconds, and that last step to create the shortcut takes several minutes. Not sure if that's related or what, just noting everything weird. Since I was getting BSOD's with the drivers from Armory Crate and the support page both, I decided to just let windows update run until there was nothing left, then I used iobit's driver booster to install/update any remaining drivers.

This so far has been the most stable iteration of the build. I have had one BSOD occur about an hour of using windows on the new install, same ntoskrnl.exe+4295a9 as before, but i left it running all night and it hasn't happened again. It hurts my soul to drop $3k to build a new rig, and every minute using it is spent anticipating the next BSOD. It seems like there may be some issue using the m.2 slots, but the drives check out ok. Any advice/help would be much appreciated.

Minidumps

[thekingswolf]

ok so maybe m.2 slots aren't it, got another BSOD right after posting this on the SSD. I'm about to go insane trying to figure out what hardware/driver is causing this.

051122-11656-01.dmp 5/11/2022 12:18:19 PM SYSTEM_SERVICE_EXCEPTION 0x0000003b 00000000`c0000005 fffff801`78327fff ffffe408`b7f86b50 00000000`00000000 ntoskrnl.exe ntoskrnl.exe+416b40 NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 10.0.22000.675 (WinBuild.160101.0800) x64 ntoskrnl.exe+416b40 C:\Windows\Minidump\051122-11656-01.dmp 24 15 22000 2,034,020 5/11/2022 12:19:26 PM

[DragDay7]

Hey there,

Are there any errors in event logs?

Things I would try (one by one):
1. disconnect everything except keyboard, mouse, monitor and ethernet (get rid of front panel hubs, additional drives)
2. see in bios if voltages and temperatures are fine
3. don't install any software, do drivers by yourself
4. set default bios settings, do not set xmp profile
5. try just one ram stick on A2 socket, try second one
6. try lowering down used pcie/m2 gen to 3rd
7. try disabled "pcie native power management" (advanced -> platform misc configuration)
8. try enabled "pcie native power management" and "native aspm"
9. carefully reseat cpu

Good luck man

[Silent Scone@ROG]

As above, it's easy to sound like a broken record, however, troubleshooting 101 always remove any overclocking. This includes XMP. Thread exception stop codes can quite easily be instability.

[thekingswolf]

no overclocking, BIOS is all set to optimized defaults. I was considering underclocking the gpu

[xeromist]

no overclocking, BIOS is all set to optimized defaults. I was considering underclocking the gpu

So what speed is the memory actually running at?

[thekingswolf]

So what speed is the memory actually running at?

[Silent Scone@ROG]

Given there's a large number of different stop codes, I'd consider removing the CPU and reseating. Check for any bent pins and reseat the cooler whilst checking for even mounting pressure. Also check all power connectors modular/extensions are connected correctly.