Can't render in GTX 1650S in latest stable version

Well, as the name says, for some reason besides everything working fine for the past weeks, suddenly I can’t render something without blender crashing (with an error), just closing suddenly, or straight up restarting my GPU or making my screen go fully black.

I thought that maybe it was because of the raising temperatures, but I check the temps of my components while rendering and they are on their usual range.

I tried updating the drivers, reverting the drivers to a previous version, using both the gaming and studio drivers, nothing. I tried using Vulkan and OpenGL, both crash.

And yes, the rendered preview is fine, with normal temps, running at it’s usual speed.

I’ve been having some weird issues these past few weeks but they were normally solved with just a restart of my PC or a manual restart of my GPU, but now im just insanely lost, and I don’t know if there’s anything that I can do or if my GPU is just falling apart.

Also I can’t get a crash log to come up again, but thankfully beacuse of my search history it crashed because of a

EXCEPTION_ACCESS_VIOLATION

and

VCRUNTIME140.dll :“Symbols not available”

seemed to be the only error

Hi there,
Does Blender give any errors before it crashes. If you run Blender from the command line it will usually output any errors to the console.

EDIT: Just seen your other post, I missed that as I responded via email.

Hi there again,
Whilst rendering if you monitor the system memory and VRAM usage, does it get close to the GTX 1650s 4gb of VRAM.

It might be that the card just doesn’t have enough VRAM to render the scene.

Yeah it may be that, but I doubt it since that usually gives the out of memory error, besides I tried rendering past scenes that rendered without issue and they still crash.

maybe you need to update your microsoft runtime library.It could be that the driver has not updated it during install.usually this librarys gets updated with the windows update but you can search and install for the librarys and update manualy iirc.

Updating Microsoft runtimes will probably solve that error, but from my experience EXCEPTION_ACCESS_VIOLATION usually points towards a VRAM issue.

I see, I will try that

Well tried that first a “non-violent” crash where blender just closed, tried once again and my screen turned black… was checking the task manager performance tab the whole time and it had a sudden spike and that’s it, I really don’t know anymore.

Was checking the VRAM usage as well but it didn’t seem to go over 3.8/4GB so idk at this point

image

3.8/4 is still quite high, since that leaves just 200mb. It likely tries to fill up more and then just goes up to 4gb before coming back down when it crashes.

If it is VRAM / GPU related temporarily switching to CPU rendering would solve the isuse.

Again makes sense, but I’m still confused since once again everything was going fine before and I tried previous scenes that I could render without issue and they still crashed.

Also should probably mention, I’m rendering on EEVEE, not cycles if that helps for anything

OK, I assumed you where on Cycles! Are your scenes large and resource heavy, such as loads of high quality image texutres?

Also looking at % usage of stuff on task manager, it actually seems like it’s only or mainly using the CPU? It’s at a constant 100% while the GPU is barely doing anything

The scene isn’t that heavy and isn’t too different to the ones I usually do, it’s just 2 characters and a normal enviroment, none of the textures should exceed 4096x4096 as far as I remember, and most are 2048x2048 or lower

I genuinely just don’t know anymore im gonna try reverting to the previous stable version

Did this, and unsurprisingly it didn’t work

Atleast I managed to get a crash log this time.

# Blender 5.1.1, Commit date: 2026-04-13 14:35, Hash b70da489d7f4

# backtrace
Exception Record:

ExceptionCode         : Microsoft C++ Exception (0xe06d7363)
Exception Address     : 0x00007FFE13445369
Exception Module      : KERNELBASE.dll
Exception Flags       : 0x00000001
Exception Parameters  : 0x4
	Parameters[0] : 0x0000000019930520
	Parameters[1] : 0x000000A4B5DFCEF0
	Parameters[2] : 0x00007FF7786A51A8
	Parameters[3] : 0x00007FF772C80000


Stack trace:
KERNELBASE.dll      :0x00007FFE13445300  RaiseException
MSVCP140.dll        :0x00007FFDBBCD2550  __ExceptionPtrRethrow
tbb12.dll           :0x00007FFDA00C8C60  tbb::detail::r1::current_context
tbb12.dll           :0x00007FFDA00C8C60  tbb::detail::r1::current_context
tbb12.dll           :0x00007FFDA00C8C60  tbb::detail::r1::current_context
blender.exe         :0x00007FF7732F2C0D  Symbols not available


Loaded Modules :

# Python backtrace

Is that whilst it’s still starting the render, such as loading materials. Keep an eye on GPU utilisation, VRAM usage, and CPU utilisation.

If it crashes whilst the GPU usage is low but the VRAM usage is high or increasing, it is an issue that happens whilst it is loading materials or loading the scene into VRAM.

Based on that crash log, I would try using a fresh install of Blender with no configuration changes or addons.

Just run the factory startup .exe; no reason to completely reinstall.

Tried, rendered it, screen went black, GPU reset, and Blender crashed I don’t even know anymore I probably need to revert to a specific drivers version or something