Home / Community / Hardware / PC Builds / RTX 4070 Drivers Crashing

RTX 4070 Drivers Crashing

Posted: 2 days ago Category: PC Builds Tags: RTX 4070, Drivers, Crashing, Troubleshooting

Hey everyone,

I'm experiencing some frustrating issues with my new RTX 4070. For the past week, my graphics drivers have been crashing intermittently, leading to black screens and game stutters. This happens randomly, sometimes during intense gaming sessions, but also sometimes when just browsing the web or watching videos.

Here's a brief rundown of my system:

  • CPU: Intel Core i7-13700K
  • Motherboard: ASUS ROG STRIX Z790-F
  • RAM: 32GB DDR5 6000MHz
  • GPU: NVIDIA GeForce RTX 4070 (MSI Ventus 3X OC)
  • PSU: Corsair RM850x (2021)
  • Storage: 1TB NVMe SSD
  • OS: Windows 11 Pro (Latest Updates)

I've already tried several troubleshooting steps:

  1. Clean Driver Installation: I've used DDU (Display Driver Uninstaller) in Safe Mode to completely remove old drivers and then installed the latest NVIDIA drivers (both the Studio and Game Ready versions) multiple times.
  2. Different Driver Versions: I've tried rolling back to older drivers, but the issue persists.
  3. Power Management: I've set the NVIDIA Control Panel power management mode to "Prefer Maximum Performance."
  4. Overclocking: I'm not currently overclocking my GPU or CPU. The GPU is running at stock speeds.
  5. Temperatures: Monitoring tools show GPU temperatures are well within limits (usually under 70°C during load).
  6. System Stability: Ran Prime95 and MemTest86, both passed without errors.
  7. Reseating GPU: I've reseated the graphics card and checked all power connections.

The crashing is really impacting my experience. Has anyone else encountered this with the RTX 4070, or have any suggestions on what else I could try? Is it possible the card itself is faulty?

Any help would be greatly appreciated!

Thanks,
John

Replies

SK
Sarah_K_Tech Posted: 1 day ago

Hey John, that sounds super frustrating! It's a common issue with new hardware sometimes. Let's try a couple more things before we jump to conclusions about the card itself.

1. Check Windows Event Viewer: When the crash happens, immediately check the Windows Event Viewer (Application and System logs) for any critical errors related to `nvlddmkm` or display driver failures. This might give us a more specific error code.

2. Test with a Different PSU Cable: Even though your PSU is good, a faulty PCIe power cable or a loose connection on the PSU side could cause intermittent power delivery issues. If you have a spare cable, try swapping it out.

3. Underclocking/Undervolting: As a test, try slightly underclocking your RTX 4070 using MSI Afterburner. Sometimes, even stock cards can be a bit unstable. A small reduction in clock speed or voltage might stabilize it. If this works, it points towards a power or stability issue.

4. Test in another PCIe Slot: If your motherboard has another PCIe x16 slot, try installing the GPU there to rule out any motherboard slot issues.

Let us know what you find!

MP
MikeP_Builder Posted: 18 hours ago

Yeah, that `nvlddmkm` error can be a real headache. Sarah's suggestions are solid.

One thing that sometimes helps is disabling **Hardware-accelerated GPU scheduling** in Windows settings. It's under Graphics settings. Sometimes newer drivers have issues with this feature enabled.

Also, make sure your VBS (Virtualization-based Security) is not interfering. You can check this by typing `msinfo32` in the run dialog. Look for "Virtualization-based Security" and if it's running, you might consider disabling it temporarily for testing (though be mindful of security implications).

Hope you get it sorted!

JD
JohnDoe_Builds Posted: 8 hours ago

Thanks for the quick replies, Sarah and Mike!

I've just checked the Event Viewer, and I'm seeing a lot of `nvlddmkm` errors logged around the times of the crashes. This is good to know!

I'll try swapping the PSU cable tomorrow, and I'll also test with a slight underclock using Afterburner. Disabling Hardware-accelerated GPU scheduling is also on my list.

I'll report back with any findings!

SK
Sarah_K_Tech Posted: 3 hours ago

Great! Seeing the `nvlddmkm` errors is a strong indicator the issue is indeed with the NVIDIA driver or GPU communication. Swapping the cable and testing underclocking are definitely the next best steps.

If none of those resolve it, and if you have the means, testing the GPU in another system would be the ultimate way to confirm if it's the card or something specific to your setup. But let's exhaust these software/connection options first!

Post a Reply