Random BSODs while gaming (0x50, 0x7F, 0xA, 0xC2), Core Ultra 9 285K / B860M, possible memory corruption

Question

Random BSODs while gaming (0x50, 0x7F, 0xA, 0xC2), Core Ultra 9 285K / B860M, possible memory corruption

Dibyendu Sinha 20

System

CPU: Intel Core Ultra 9 285K
Motherboard: GIGABYTE B860M AORUS ELITE WIFI6E, BIOS F11 (updated from F10 shortly before the crashes began)
RAM: 32 GB DDR5 (2 × 16 GB), running at 4800 MT/s (XMP off), slots 2 of 4
GPU: ZOTAC RTX 5070 Ti AMP Extreme INFINITY, NVIDIA driver 617.42 (clean install)
Cooler: MSI MAG CoreLiquid A13 360
PSU: 850 Watt 80+ Gold
OS: Windows 11 Pro, build 26100.9457, Memory Integrity on

Problem

Since Oct 29, 2026, I’ve had 4 BSODs in about 3 days, all while gaming (PUBG, Phantom Blade Zero, Forza Horizon 6). Every crash has a different bugcheck code and a different faulting module. Crashes 2–4 have only Microsoft modules on the stack.

#DateBugcheckFaulting moduleProcess1Oct 290x50 PAGE_FAULT_IN_NONPAGED_AREAnavagio.sys (PUBG)TslGame.exe----------------------------------------1Oct 290x50 PAGE_FAULT_IN_NONPAGED_AREAnavagio.sys (PUBG)TslGame.exe2Oct 290x7F UNEXPECTED_KERNEL_MODE_TRAP (arg1 = 8, double fault, bad stack pointer 0xffff)FLTMGR.SYSPBZ.exe3Oct 300xA IRQL_NOT_LESS_OR_EQUAL (write to 0x40)ntkrnlmp.exe (KiTryUnwaitThread)MemCompression4Nov 10xC2 BAD_POOL_CALLER (arg1 = 0x99, freeing address 0x0000000001000000)Ntfs.sys (NtfsDeallocateCompressionBuffer)dwm.exeIn crash 4, the freed address is zero with a single bit set, which looks like a bit flip.

Changes made shortly before the crashes started

BIOS updated F10 → F11 through GIGABYTE Control Center
NVIDIA driver 621.01 installed
GIGABYTE EasyTune / Performance Library / GBTECService installed (Oct 27)

What I’ve already done

Rolled the NVIDIA driver back to 617.42 (clean install). Crashes 3 and 4 happened after the rollback.
Uninstalled GIGABYTE Control Center, EasyTune, Performance Library, GBTECService and AI Snatch. Crash 4 happened after this.
XMP disabled; RAM runs at JEDEC 4800 MT/s.
sfc /scannow: no integrity violations.
fltmc: only Microsoft minifilters loaded (WdFilter, bindflt, gameflt, CldFlt, etc.).

Planned next: roll the BIOS back to F10, and run MemTest86 (4 passes).

Questions

Does the pattern across these four dumps point to faulty RAM, the CPU’s memory controller, or the F11 BIOS?
Would you recommend running Driver Verifier, and if so, with which settings?
Anything else I should test before replacing hardware?

Dump files: [OneDrive/Google Drive link to the zipped .dmp files]

!analyze -v outputs (Analysis/Hypervisor key-value blocks removed)

Crash 1 (102926-14234-01.dmp):

#DateBugcheckFaulting moduleProcess1Oct 290x50 PAGE_FAULT_IN_NONPAGED_AREAnavagio.sys (PUBG)TslGame.exe2Oct 290x7F UNEXPECTED_KERNEL_MODE_TRAP (arg1 = 8, double fault, bad stack pointer 0xffff)FLTMGR.SYSPBZ.exe3Oct 300xA IRQL_NOT_LESS_OR_EQUAL (write to 0x40)ntkrnlmp.exe (KiTryUnwaitThread)MemCompression4Nov 10xC2 BAD_POOL_CALLER (arg1 = 0x99, freeing address 0x0000000001000000)Ntfs.sys (NtfsDeallocateCompressionBuffer)dwm.exeIn crash 4, the freed address is zero with a single bit set, which looks like a bit flip.

Changes made shortly before the crashes started

BIOS updated F10 → F11 through GIGABYTE Control Center
NVIDIA driver 621.01 installed
GIGABYTE EasyTune / Performance Library / GBTECService installed (Oct 27)

What I’ve already done

Rolled the NVIDIA driver back to 617.42 (clean install). Crashes 3 and 4 happened after the rollback.
Uninstalled GIGABYTE Control Center, EasyTune, Performance Library, GBTECService and AI Snatch. Crash 4 happened after this.
XMP disabled; RAM runs at JEDEC 4800 MT/s.
sfc /scannow: no integrity violations.
fltmc: only Microsoft minifilters loaded (WdFilter, bindflt, gameflt, CldFlt, etc.).

Planned next: roll the BIOS back to F10, and run MemTest86 (4 passes).

Questions

Does the pattern across these four dumps point to faulty RAM, the CPU’s memory controller, or the F11 BIOS?
Would you recommend running Driver Verifier, and if so, with which settings?
Anything else I should test before replacing hardware?

Dump files: [OneDrive/Google Drive link to the zipped .dmp files]

!analyze -v outputs (Analysis/Hypervisor key-value blocks removed)

Crash 1 (102926-14234-01.dmp):

PAGE_FAULT_IN_NONPAGED_AREA (50)
BUGCHECK_CODE:  50
BUGCHECK_P1: ffffffffff45c12e
BUGCHECK_P2: 0
BUGCHECK_P3: fffff80095facbb9
BUGCHECK_P4: 2
Failure.Bucket: AV_navagio!unknown_function
Faulting.IP.Type: Paged
FAULTING_THREAD:  ffffac841c17e080
ERROR_CODE: (NTSTATUS) 0xc0000005
READ_ADDRESS: ffffffffff45c12e
MM_INTERNAL_CODE:  2
PROCESS_NAME:  TslGame.exe
IP_IN_PAGED_CODE:
navagio+45cbb9
fffff800`95facbb9 45334141        xor     r8d,dword ptr [r9+41h]

STACK_TEXT:
ffff960f`ddeb6d18 fffff800`debc8a75     : 00000000`00000050 ffffffff`ff45c12e 00000000`00000000 ffff960f`ddeb6f80 : nt!KeBugCheckEx
ffff960f`ddeb6d20 fffff800`dea961e6     : ffffac84`0ba434c0 ffff8000`00000000 ffffffff`ff45c12e 0000007f`fffffff8 : nt!MiSystemFault+0x661
ffff960f`ddeb6e10 fffff800`deebd8cb     : fffff800`9635266b ffff960f`ddeb7000 ffff960f`ddeb73a4 00001f80`00a2008a : nt!MmAccessFault+0x646
ffff960f`ddeb6f80 fffff800`95facbb9     : 00000000`00000000 fffff800`96599ae3 ffffac84`1136f018 fffff800`dea17da0 : nt!KiPageFault+0x38b
ffff960f`ddeb7110 00000000`00000000     : fffff800`96599ae3 ffffac84`1136f018 fffff800`dea17da0 00000000`00040202 : navagio+0x45cbb9

SYMBOL_NAME:  navagio+45cbb9
MODULE_NAME: navagio
IMAGE_NAME:  navagio.sys
FAILURE_BUCKET_ID:  AV_navagio!unknown_function

Crash 2 (102926-14406-01.dmp):

UNEXPECTED_KERNEL_MODE_TRAP (7f)
Arg1: 0000000000000008, EXCEPTION_DOUBLE_FAULT
Arg2: ffffa780bc1bde70
Arg3: 000000000000ffff
Arg4: fffff8040c5771dc
Bugcheck.Code.LegacyAPI: 0x1000007f
Failure.Bucket: 0x7f_8_FLTMGR!FltpTraceIOStatusOnFailure
Stack.Pointer: Invalid
FAULTING_THREAD:  ffffd005ddd57240
BAD_STACK_POINTER:  000000000000ffff
PROCESS_NAME:  PBZ.exe
UNALIGNED_STACK_POINTER:  000000000000ffff

STACK_TEXT:
00000000`0000ffff 00000000`00000000     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : FLTMGR!FltpTraceIOStatusOnFailure+0x4c

SYMBOL_NAME:  FLTMGR!FltpTraceIOStatusOnFailure+4c
MODULE_NAME: FLTMGR
IMAGE_NAME:  FLTMGR.SYS
IMAGE_VERSION:  10.0.26100.9278
FAILURE_BUCKET_ID:  0x7f_8_FLTMGR!FltpTraceIOStatusOnFailure

Crash 3 (103026-13312-01.dmp):

IRQL_NOT_LESS_OR_EQUAL (a)
Arg1: 0000000000000040, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000001, write operation
Arg4: fffff803c7fa8803, address which referenced memory
Failure.Bucket: AV_nt!KiTryUnwaitThread
Faulting.IP.Type: Paged
FAULTING_THREAD:  ffff838fdcdf8080
WRITE_ADDRESS: 0000000000000040
PROCESS_NAME:  MemCompression
IP_IN_PAGED_CODE:
nt!KiTryUnwaitThread+33
fffff803`c7fa8803 f0480fba6b4000  lock bts qword ptr [rbx+40h],0

STACK_TEXT:
ffffe00a`438f6ee8 fffff803`c82c26e9     : 00000000`0000000a 00000000`00000040 00000000`00000002 00000000`00000001 : nt!KeBugCheckEx
ffffe00a`438f6ef0 fffff803`c82bd9a8     : 00000000`00000000 00000000`00000000 00000000`00000000 ffff838f`c6763c40 : nt!KiBugCheckDispatch+0x69
ffffe00a`438f7030 fffff803`c7fa8803     : ffffe00a`438f71e8 00000000`00000000 ffff838f`dd329ed0 00000000`00000000 : nt!KiPageFault+0x468
ffffe00a`438f71c0 fffff803`c7e5bab5     : 00000000`00000000 00000000`00000000 ffff838f`d6a09b10 00000000`00000001 : nt!KiTryUnwaitThread+0x33
ffffe00a`438f7220 fffff803`c7e33c2b     : ffff838f`00000000 00000000`00000000 00000000`00000000 00000000`00000002 : nt!KeSetEvent+0x6d5
ffffe00a`438f72b0 fffff803`c7f1392f     : 00000000`00000001 ffff838f`dd329ed0 ffff838f`d6a09d08 ffff838f`d6a08000 : nt!SMKM_STORE<SM_TRAITS>::SmStWorkItemQueue+0x127
ffffe00a`438f7300 fffff803`c7fd50f9     : 00000000`00000000 00000000`00000002 ffff838f`ffffffff fffff803`00000000 : nt!ST_STORE<SM_TRAITS>::StDmLazyWorkItemQueue+0x4b
ffffe00a`438f7330 fffff803`c7fd4fbb     : ffff838f`d6a08000 ffffe00a`438f74ff ffff838f`e8968000 ffff838f`d6a09d08 : nt!SMKM_STORE<SM_TRAITS>::SmStDirectReadIssue+0xb5
ffffe00a`438f73b0 fffff803`c7ebad87     : ffff838f`dcdf80fd ffff838f`dd329ed0 00000000`00000000 00000000`00000002 : nt!SMKM_STORE<SM_TRAITS>::SmStDirectReadCallout+0x2b
ffffe00a`438f73e0 fffff803`c7ebae4d     : fffff803`c7fd4f90 ffffe00a`438f74c0 00000000`00000002 ffff838f`c67636b8 : nt!KeExpandKernelStackAndCalloutInternal+0xb7
ffffe00a`438f7450 fffff803`c7fd376f     : 00000000`00000002 00000000`00000000 00000000`00000000 ffff838f`e329e248 : nt!KeExpandKernelStackAndCalloutEx+0x1d
ffffe00a`438f7490 fffff803`c7e33cb8     : 00000000`00000001 ffff838f`ef6ff000 ffff838f`d6a08000 ffff838f`e329e150 : nt!SMKM_STORE<SM_TRAITS>::SmStDirectRead+0xcf
ffffe00a`438f7560 fffff803`c7e33a75     : 00000000`00000000 ffffe00a`438f7639 00000000`00000000 ffff838f`40000010 : nt!SMKM_STORE<SM_TRAITS>::SmStWorkItemQueue+0x1b4
ffffe00a`438f75b0 fffff803`c7e336f4     : 00000000`00000000 00000000`00000000 ffffe00a`438f7639 ffff838f`dd329ed0 : nt!SmWorkItemQueue+0x9
ffffe00a`438f75e0 fffff803`c7fcda93     : ffff838f`e329e150 00000000`0000000c ffff838f`00000000 ffff838f`dd329ed0 : nt!SMKM_STORE_MGR<SM_TRAITS>::SmIoCtxQueueWork+0x1e4
ffffe00a`438f76a0 fffff803`c7fcd952     : 00000000`00000002 ffffe00a`438f7760 ffffe00a`00000002 fffff803`00000000 : nt!SMKM_STORE_MGR<SM_TRAITS>::SmPageRead+0x133
ffffe00a`438f7720 fffff803`c7f709be     : ffff838f`0001af0d ffff838f`d7bba580 00000000`00000001 fffff803`c7e94491 : nt!SmPageRead+0x22
ffffe00a`438f7760 fffff803`c7f6f905     : 00000000`00000002 00000000`00000001 00000000`00000000 00000000`00000004 : nt!MiIssueHardFaultIo+0x92
ffffe00a`438f77b0 fffff803`c7e96014     : ffff8000`00000000 00000159`0a4bb8a0 00000000`c0033333 fffff803`c847147e : nt!MiIssueHardFault+0x251
ffffe00a`438f78b0 fffff803`c82bd8cb     : ffff838f`dcdf8080 00000000`00000000 00000000`00000000 0000002c`b32ff918 : nt!MmAccessFault+0x474
ffffe00a`438f7a20 00007ffc`85826863     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiPageFault+0x38b
0000002c`b32ff598 00000000`00000000     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x00007ffc`85826863

SYMBOL_NAME:  nt!KiTryUnwaitThread+33
MODULE_NAME: nt
IMAGE_NAME:  ntkrnlmp.exe
IMAGE_VERSION:  10.0.26100.9457
FAILURE_BUCKET_ID:  AV_nt!KiTryUnwaitThread

Crash 4 (110126-11640-01.dmp):

BAD_POOL_CALLER (c2)
Arg1: 0000000000000099, Attempt to free pool with invalid address (or corruption in pool header)
Arg2: 0000000001000000, Address being freed
Arg3: 0000000000000000
Arg4: 0000000000000000
Failure.Bucket: 0xc2_99_Ntfs!NtfsDeallocateCompressionBuffer
Faulting.IP.Type: Paged
FAULTING_THREAD:  ffffba0cad019080
PROCESS_NAME:  dwm.exe
IP_IN_PAGED_CODE:
nt!ExFreePoolWithTag+df
fffff805`a318094f cc              int     3

STACK_TEXT:
fffff00c`365befd8 fffff805`a318094f     : 00000000`000000c2 00000000`00000099 00000000`01000000 00000000`00000000 : nt!KeBugCheckEx
fffff00c`365befe0 fffff805`35bc6ce1     : 00000000`01000000 ffffba0c`8a6981b0 00000000`00401000 00000000`c0000001 : nt!ExFreePoolWithTag+0xdf
fffff00c`365bf0e0 fffff805`35dce439     : ffffba0c`ab9f6cc8 00000000`00001000 00000000`00000000 00000000`00000000 : Ntfs!NtfsDeallocateCompressionBuffer+0x481
fffff00c`365bf180 fffff805`35bc015e     : fffff00c`365bf500 ffffba0c`ab9f6cc8 ffff9504`c2c3a710 ffffba0c`ab9f6cc8 : Ntfs!NtfsNonCachedIo+0xea9
fffff00c`365bf320 fffff805`35bbf275     : ffffba0c`ab9f6cc8 ffffba0c`9f276850 fffff00c`365bf508 ffffba0c`b2f05d10 : Ntfs!NtfsCommonRead+0x7ae
fffff00c`365bf470 fffff805`a28b2254     : ffffba0c`9d9d79a0 ffffba0c`9f276850 ffffba0c`9d9d7a08 00000000`00000000 : Ntfs!NtfsFsdRead+0x285
fffff00c`365bf5b0 fffff805`3463b45d     : ffffba0c`95f685d0 ffffba0c`9d9d79a0 00000000`00000000 ffffba0c`aedeb480 : nt!IofCallDriver+0x84
fffff00c`365bf5f0 fffff805`3463c758     : fffff00c`365bf680 00000000`00000000 ffff8880`19146100 ffff8000`00000000 : FLTMGR!FltpLegacyProcessingAfterPreCallbacksCompleted+0x16d
fffff00c`365bf660 fffff805`a28b2254     : ffffba0c`aedeb080 00000001`00000000 00000000`00000020 00000000`00000004 : FLTMGR!FltpDispatch+0xc8
fffff00c`365bf6c0 fffff805`a28b1bd6     : ffffba0c`aedeb080 ffffba0c`b2f05e00 00000001`00000000 00000000`00000000 : nt!IofCallDriver+0x84
fffff00c`365bf700 fffff805`a2970a7b     : 00000000`00000001 fffff00c`365bf840 ffffba0c`b2f05d50 ffffba0c`b2f05d10 : nt!IoPageReadEx+0x2c6
fffff00c`365bf760 fffff805`a296f905     : 00000000`00000003 00000000`00000001 00000000`00000000 00000000`00000004 : nt!MiIssueHardFaultIo+0x14f
fffff00c`365bf7b0 fffff805`a2896014     : ffff8000`00000000 00007fff`b0650856 00000000`c0033333 ffff9504`d02267e0 : nt!MiIssueHardFault+0x251
fffff00c`365bf8b0 fffff805`a2cbd8cb     : ffffba0c`ad019080 000001a4`19058860 00000000`00000010 00000000`00000000 : nt!MmAccessFault+0x474
fffff00c`365bfa20 00007fff`b6e9429c     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiPageFault+0x38b
00000017`83afc680 00000000`00000000     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x00007fff`b6e9429c

SYMBOL_NAME:  Ntfs!NtfsDeallocateCompressionBuffer+481
MODULE_NAME: Ntfs
IMAGE_NAME:  Ntfs.sys
IMAGE_VERSION:  10.0.26100.9444
FAILURE_BUCKET_ID:  0xc2_99_Ntfs!NtfsDeallocateCompressionBuffer

0 comments

Answer accepted by question author

0 additional answers

Your answer

Answer 1

Kai-H 30,315 Microsoft External Staff Moderator

Hi, Dibyendu Sinha

Four different crash codes inside three days, mostly in Windows' own files, usually means something is quietly corrupting memory under load. Since it started right after the F11 update, I'd look at what that BIOS might have changed on its own first.

Here are some suggestions you can try:

Check that your RAM is really running at stock. XMP being off doesn't always mean that. Gigabyte BIOSes have a feature called DDR5 Auto Booster that can push standard RAM above its normal speed even with XMP off. There are also High Bandwidth and Low Latency options that tighten memory settings. A BIOS update can turn these back on. In the BIOS, go to the Tweaker page and set all three to Disabled. Then open CPU-Z, go to the Memory tab and check that the DRAM Frequency shows about 2400 MHz.

Stress test inside Windows. MemTest86 runs while the CPU is cool and mostly idle. Gaming crashes often only show up when the CPU and memory are both hot and busy. Run the OCCT Memory test or the y-cruncher stress test for about an hour. You can also run the free Intel Processor Diagnostic Tool. If any of these fail while the RAM is confirmed at 4800, that points more to the CPU or board than to normal RAM settings.

Do a clean reinstall of the PUBG anti-cheat (navagio.sys). Uninstall PUBG. Then delete these folders: %localappdata%\TslGame, %localappdata%\WELLBIA, C:\Program Files\AntiCheatExpert and C:\Program Files\Common Files\PUBG. After that, clear the Steam download cache and reinstall the game.

If the OCCT or y-cruncher test fails at stock settings with Auto Booster off, the problem is most likely hardware. At that point it's worth starting an RMA with Gigabyte or Intel.

Dibyendu Sinha 20 Reputation points

2026-11-01T17:52:54.93+00:00

Hey Kai-H, thanks a lot for the reply, it really helped. You were right about the BIOS memory settings. I checked and DDR5 XMP Booster, High Bandwidth and Low Latency were all turned on by default (set to Auto/Enabled), on both F11 and F10. I had no idea those were even there.

I rolled the BIOS back to F10 and turned all three of those off/Disabled. CPU-Z now shows the RAM at 2400 MHz (so 4800) with 40-40-40-77 timings, which matches the stock JEDEC profile. My sticks are actually a G.Skill 6000 kit, but I’m leaving them at 4800 for now until I’m sure everything is stable.

So far I’ve played about 2 hours on F10 (some Forza Horizon 6 and Phantom Blade Zero) without any crash, but that was before I disabled the memory settings, so I’m going to keep playing normally for the next few days and see how it goes. If it crashes again I’ll run OCCT/y-cruncher and MemTest86 like you said and post the results here.

I’ll update this thread either way. Thanks again!
Kai-H 30,315 Reputation points Microsoft External Staff Moderator

2026-11-02T06:12:16.3666667+00:00

Glad to see the improvements on your situation.

If there be any further updates, feel free to let me know.

Have a nice day!
Dibyendu Sinha 20 Reputation points

2026-11-03T20:52:25.6966667+00:00

Hey Kai-H, just wanted to give you an update. It’s been about 3 days since I rolled back to F10 and disabled DDR5 XMP Booster, High Bandwidth and Low Latency, and I haven’t had a single crash since. Before this, the PC would usually BSOD after about 1–2 hours under gaming load, so it’s a big difference.

I’ve been playing PUBG for around 4 hours a day on ultra settings, with Brave and a bunch of tabs open in the background, and everything has been completely stable. The RAM is still at stock 4800 with JEDEC timings, and I’m planning to leave it like that for now.

I’ll keep playing Phantom Blade Zero and Forza Horizon 6 over the next few days too, since most of the crashes happened in those, and I’ll post here again if anything comes back. Thanks a lot for pointing me to those memory settings. I honestly wouldn’t have found them on my own.
Kai-H 30,315 Reputation points Microsoft External Staff Moderator

2026-11-04T08:52:35.3433333+00:00

It's good to hear that.

Please let me know if you need more help in the future!