Windows 10 BSOD during rendering

My scene consists of a 400x400m ocean modifier plane with resolution set to 20 and 240 frames of cache, along with 6 instanced collection of a particle system (300 instances of a uv sphere). I have tried rendering this scene 30+ times and everytime it either crashes blender or crashes windows entirely. I need help figuring out what is causing this problem.

Shoreline.blend (3.2 MB)

I have tried to open one of the latest blue screen .dmp but can’t understand what it says about the error.

Microsoft (R) Windows Debugger Version 10.0.22549.1000 AMD64
Copyright (c) Microsoft Corporation. All rights reserved.


Loading Dump File [C:\Windows\Minidump\051922-8625-01.dmp]
Mini Kernel Dump File: Only registers and stack trace are available

Symbol search path is: srv*
Executable search path is: 
Windows 10 Kernel Version 19041 MP (16 procs) Free x64
Product: WinNt, suite: TerminalServer SingleUserTS
Edition build lab: 19041.1.amd64fre.vb_release.191206-1406
Machine Name:
Kernel base = 0xfffff807`75400000 PsLoadedModuleList = 0xfffff807`7602a2b0
Debug session time: Thu May 19 00:43:38.320 2022 (UTC - 4:00)
System Uptime: 0 days 0:55:47.906
Loading Kernel Symbols
...............................................................
................................................................
................................................................
............................................
Loading User Symbols
Loading unloaded module list
..........
For analysis of this file, run !analyze -v
nt!KeBugCheckEx:
fffff807`757f7d60 48894c2408      mov     qword ptr [rsp+8],rcx ss:0018:fffffd84`e25999f0=000000000000000a
4: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

IRQL_NOT_LESS_OR_EQUAL (a)
An attempt was made to access a pageable (or completely invalid) address at an
interrupt request level (IRQL) that is too high.  This is usually
caused by drivers using improper addresses.
If a kernel debugger is available get the stack backtrace.
Arguments:
Arg1: ffffffff8b481042, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000001, bitfield :
	bit 0 : value 0 = read operation, 1 = write operation
	bit 3 : value 0 = not an execute operation, 1 = execute operation (only on chips which support this level of status)
Arg4: fffff807756db3b6, address which referenced memory

Debugging Details:
------------------

*** WARNING: Unable to verify checksum for win32k.sys

KEY_VALUES_STRING: 1

    Key  : Analysis.CPU.mSec
    Value: 2436

    Key  : Analysis.DebugAnalysisManager
    Value: Create

    Key  : Analysis.Elapsed.mSec
    Value: 7802

    Key  : Analysis.Init.CPU.mSec
    Value: 530

    Key  : Analysis.Init.Elapsed.mSec
    Value: 19142

    Key  : Analysis.Memory.CommitPeak.Mb
    Value: 97

    Key  : WER.OS.Branch
    Value: vb_release

    Key  : WER.OS.Timestamp
    Value: 2019-12-06T14:06:00Z

    Key  : WER.OS.Version
    Value: 10.0.19041.1


FILE_IN_CAB:  051922-8625-01.dmp

BUGCHECK_CODE:  a

BUGCHECK_P1: ffffffff8b481042

BUGCHECK_P2: 2

BUGCHECK_P3: 1

BUGCHECK_P4: fffff807756db3b6

WRITE_ADDRESS: fffff807760fb390: Unable to get MiVisibleState
Unable to get NonPagedPoolStart
Unable to get NonPagedPoolEnd
Unable to get PagedPoolStart
Unable to get PagedPoolEnd
unable to get nt!MmSpecialPagesInUse
 ffffffff8b481042 

BLACKBOXBSD: 1 (!blackboxbsd)


BLACKBOXNTFS: 1 (!blackboxntfs)


BLACKBOXPNP: 1 (!blackboxpnp)


BLACKBOXWINLOGON: 1

CUSTOMER_CRASH_COUNT:  1

PROCESS_NAME:  System

TRAP_FRAME:  fffffd84e2599b30 -- (.trap 0xfffffd84e2599b30)
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=fffffd84e2599d18 rbx=0000000000000000 rcx=ffffba8092500180
rdx=0000000000000000 rsi=0000000000000000 rdi=0000000000000000
rip=fffff807756db3b6 rsp=fffffd84e2599cc0 rbp=fffffd84e2599d59
 r8=000000000000fe00  r9=0000000000000000 r10=fffff807760e8c00
r11=0000000000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up ei ng nz na po nc
nt!KiCheckForThreadDispatch+0x12:
fffff807`756db3b6 488b4110        mov     rax,qword ptr [rcx+10h] ds:ffffba80`92500190=0000000000000000
Resetting default scope

STACK_TEXT:  
fffffd84`e25999e8 fffff807`75809c69     : 00000000`0000000a ffffffff`8b481042 00000000`00000002 00000000`00000001 : nt!KeBugCheckEx
fffffd84`e25999f0 fffff807`75805f69     : 00000000`0000000b fffffd84`e2599b5c fffffd84`e2599ba0 00000000`00000000 : nt!KiBugCheckDispatch+0x69
fffffd84`e2599b30 fffff807`756db3b6     : ffffba80`925088f0 ffffba80`92500180 ffffe20e`fa60f6d0 00000000`00000000 : nt!KiPageFault+0x469
fffffd84`e2599cc0 fffffd84`e2599d59     : fffffd84`e2599d58 fffff807`00000000 ffffba80`92500180 fffff807`7591a23e : nt!KiCheckForThreadDispatch+0x12
fffffd84`e2599d00 fffffd84`e2599d58     : fffff807`00000000 ffffba80`92500180 fffff807`7591a23e 00000000`0000ffff : 0xfffffd84`e2599d59
fffffd84`e2599d08 fffff807`00000000     : ffffba80`92500180 fffff807`7591a23e 00000000`0000ffff fffffd84`e2599d90 : 0xfffffd84`e2599d58
fffffd84`e2599d10 ffffba80`92500180     : fffff807`7591a23e 00000000`0000ffff fffffd84`e2599d90 00000000`00000000 : 0xfffff807`00000000
fffffd84`e2599d18 fffff807`7591a23e     : 00000000`0000ffff fffffd84`e2599d90 00000000`00000000 00000000`00000000 : 0xffffba80`92500180
fffffd84`e2599d20 ffffba80`92500180     : 00000000`00000010 00000000`00000000 ffffe20e`fa60f720 00000000`00000000 : nt!KiConvertDynamicHeteroPolicy+0x1e
fffffd84`e2599d50 00000000`00000010     : 00000000`00000000 ffffe20e`fa60f720 00000000`00000000 00000000`00000000 : 0xffffba80`92500180
fffffd84`e2599d58 00000000`00000000     : ffffe20e`fa60f720 00000000`00000000 00000000`00000000 fffff780`00000320 : 0x10


SYMBOL_NAME:  nt!KiCheckForThreadDispatch+12

MODULE_NAME: nt

IMAGE_NAME:  ntkrnlmp.exe

IMAGE_VERSION:  10.0.19041.1706

STACK_COMMAND:  .cxr; .ecxr ; kb

BUCKET_ID_FUNC_OFFSET:  12

FAILURE_BUCKET_ID:  AV_nt!KiCheckForThreadDispatch

OS_VERSION:  10.0.19041.1

BUILDLAB_STR:  vb_release

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {c98a5eda-5400-cb5a-001c-66777c4a4a88}

Followup:     MachineOwner
---------

PC components:

		Operating System
			Windows 10 Pro 64-bit
		CPU
			AMD Ryzen 7 5800X 8-Core Processor (-30 underclock on all cores)
		RAM
			32.0GB Dual-Channel Unknown @ 1599MHz (22-22-22-53)
		Motherboard
			Gigabyte Technology Co. Ltd. X570 AORUS ELITE WIFI (AM4)	32 °C
		Graphics
			G27QC (2560x1440@144Hz)
			4095MB NVIDIA GeForce RTX 3080 (Gigabyte)	44 °C
		Storage
			1863GB Western Digital WDC WD20EZBX-00AYRA0 (SATA (SSD))	39 °C
			931GB Samsung SSD 970 EVO Plus 1TB (Unknown (SSD))
		Optical Drives
			No optical disk drives detected
		Audio
			NVIDIA High Definition Audio

Hi 30Salt,

Tried your blend file on my machine (old Xeon 24GB ram and GTX 1060 6GB) and it ran without crashing using E-Cycles 3.1.2. Although all frames look the same (ran in 9m 72s). Even after, I adjusted the path within Compositing & output folder outputs still looks all the same. Perhaps, I maybe missed something else (I’m generally on the newbie side).

Have you checked your Blender Crash Log? Assuming Blender got the chance to create one. These can generally be found within the folder specified in User Preferences > File Paths > Temporary Files. Or within your Windows temp folder (likely C:\Users\YourUserName\AppData\Local\Temp\ ) but this may vary depending on your config (see here) for more info.

I’m no Windows expert but it looks like a kernel issue which most of the time points to a driver problem. Hopefully nothing worse. I’ve been hit by that so many times it’s not funny. I’d recommend a tool like https://www.iobit.com/en/driver-booster.php. I found it useful enough to purchase the full version. I did find the latest NVidia drivers buggy with Blender so I downgraded my video drivers to 2 versions ago and all is good now (fingers crossed). (Driver 512.15 (2020-May-09) although note that my card is much older than yours so this might not work for you).

Also, noticed that your RTX 3080 should have at least 8GB of memory not 4GB as you listed.

Good to check your system while rending with Task Manager (or equivalent) and your thermals, memory / resources. Just to make sure it’s not that. Eevee isn’t as bad as Cycles though.

Other useful links:

Cheers and good luck.

1 Like

Hi pxlpaul,

Thanks for your troubleshooting tips, I have gone through most of the steps you provided and it really helped a lot to pinpoint where the issues are.

The static render was my mistake, I did not upload the cache files for the ocean modifier or the baked .mdd files for the shrinkwrap plane that I used to drive the location and rotation of the main objects and the camera. (I baked the shrinkwrap modifier into a .mdd file so that I can simulate a delay in its animation, so objects floating in water lag 3 frames behind the waves instead of matching the surface of water perfectly.)

In terms of a crash log, thanks for letting me know where it is located as Blender didn’t leave any error message and I did not know where to find a crash log. Unfortunately, I did not find any crash logs where the .blend file is located or at .\AppData\Local\Temp\ (I have not changed the file paths for temporary files).

The thermals were great just from looking at the task manager. So I figured it might be something with the drivers.

I used the DDU tool in safe mode to uninstall all drivers (even for audio). Most drivers installed themselves correctly without any input upon reboot with the exception of the Nvidia graphics driver, which I then installed using GeForce experience (which might be the source of the problem as the latest versions could be buggy as you mentioned). Even then I was having random crashes and BSOD using blender or playing games, so I tried doing a fresh install of Windows 11. I was able to render the whole duration of the blend scene at resolution 20 without crashing. I will further test this with increased ocean modifier resolution. On the gaming side, the immediate fatal crashes I was experiencing in games like No Man’s Sky are gone, but I still got a java error and cannot launch Forge Minecraft Clients (servers ran fine).

Yesterday I tried turning the X.M.P profile on in BIOS. So far it seems to have fixed all issues, blender runs and renders without crashing and I was able to run the Minecraft client/server continuously overnight. I will test with more programs in the following days, but most likely everything is fine now.

As a side note, I was also weirded out by the speccy hardware summary, my RTX 3080 should have 10GB video memory and I don’t know why it says 4GB. It is from gigabyte which is a brand that I can’t trust fully. Is this something worth looking into?

Thanks a ton for your help!

Gigabyte is as trustworthy as any other manufacturer. If they were selling 4GB 3080s, it would be major news. Did you buy the hardware from a reputable storefront?
It’s probably just a reporting error if you’ve eliminated the blue screens now though. Check windows display settings and dxdiag.

2 Likes

If I may suggest something to consider:

Which is more likely- that Gigabyte has been bamboozling hundreds of thousands (https://www.userbenchmark.com/PCBuilder?tab=GPU&searchTerm=1458+403F, for example, lists 13,902 users of the Gigabyte RTX 3080, and that’s just users that have benchmarked it on that specific site) of users about how much VRAM it has, or that your computer is experiencing some kind of error?

1 Like

I bought the card from Canada Computers. Promise I’m not hating on Gigabyte, had some issues with their hardware in the past is all. dxdiag is super useful and I’m really thankful that you brought it up. The results are as follows:

Turns out it was a false report by speccy. Blender is now able to handle heavy scenes without crashing, and Windows 11 is now stable. Thank you all for your input! :ok_hand:

1 Like