Intermittent crashes with repeated VIDEO_TDR_FAILURE (0x116)

Question

Hey,

I’m hoping to get some input on an ongoing crash issue I’ve been dealing with for a while now. The system crashes are intermittent and unpredictable, but the crash dumps have been very consistent, which is why I’m reaching out.

The PC can run perfectly fine for long periods of time — sometimes days or even weeks — and then suddenly I’ll start getting crashes. When it happens, the screen goes black and the system resets or reboots abruptly. Sometimes Windows can’t fully write the crash dump afterward.

There doesn’t seem to be a single trigger. It’s happened during gaming, light use, and occasionally around idle. Once the instability starts, it may happen a few times before the system goes back to being stable again.

I’ve collected several kernel crash dumps over different dates, and most of them show the same thing:

BugCheck 0x116 – VIDEO_TDR_FAILURE

This has happened across multiple NVIDIA driver versions over a period of months, not just one specific driver.

Dump files I have:

111825-11718-01.dmp – 0x116

111825-12000-01.dmp – 0x116

111825-14515-01.dmp – 0x116

120125-14515-01.dmp – 0x116

There’s also one different crash:

111825-14625-01.dmp – BugCheck 0x1000007E

That one faults in Ntfs.sys with an access violation while closing a file. From what I can tell, this looks like a secondary crash after the system was already unstable, rather than a storage issue causing everything else.

In Event Viewer, I usually see volmgr Event ID 162 after the crashes, which seems to line up with the system resetting too hard to finish writing the dump.

I don’t see ongoing disk errors, NVMe controller resets, or WHEA hardware errors that would point to a clear storage or CPU fault.

So far I have:

• Tested multiple NVIDIA driver versions over several months (same crashes keep happening).
• Tried different wall outlets.
• Ran the PC plugged directly into the wall (no power boards, no extensions).
• The system has been tested in-store twice, but short stress tests and overnight idle never reproduced the issue (which makes sense given how intermittent it is).

The main thing that stands out to me is the pattern:

long periods where everything is fine

followed by sudden instability

with the same VIDEO_TDR_FAILURE showing up each time

Because this happens across driver versions and over a long timeframe, it feels less like a bad driver and more like some kind of intermittent hardware instability in the GPU path.

I’m mainly looking for a sanity check from Microsoft:

• Does this crash dump pattern look more like a hardware issue (GPU / PCIe / power delivery) than a Windows or driver bug?
• Are there any known Windows or NVIDIA driver issues that would cause persistent 0x116 crashes across many driver versions like this?
• Is there any additional Windows-side testing you’d recommend that makes sense for an issue that’s this intermittent?

Below are of the crash logs I have uploaded into a .zip and uploaded to google drive:

https://drive.google.com/file/d/1JSJhJTa8f25V29AlPqNAfy214Wodgcme/view?usp=sharing

any insight would be appreciated.

Answer 1

other specs include a 4090 that isn't mentioned in the list above

Answer 2

Thank you for providing the detailed logs and hardware information! Four of the crashes point directly to nvlddmkm.sys. The single Ntfs.sys crash is likely a secondary failure triggered by the initial GPU crash.

The problem persists across multiple driver versions and exhibits a pattern of "long-term stability → sudden crash → returns to stability." This strongly suggests a hardware issue, rather than a driver bug or a flaw in Windows itself.

I guess the key factor is the RTX 4090's high-power nature. It is extremely sensitive to power delivery. Even if the PSU's rated wattage is sufficient, factors like transient power spikes, poor contact on the 16-pin connector, or insufficient dynamic response from the power supply can all trigger a TDR timeout, sometimes even during low-load or idle states.

I recommend prioritizing the following steps to isolate the power delivery path:

1. Check the connector for any signs of burning/scorching or looseness. Use the original cable or a genuine 3*8 -pin adapter.

Test Power Supply Stability:

Use HWiNFO64 to monitor the voltage rail. If the voltage drops below 11.4V immediately preceding a crash, the PSU is likely insufficient.

If possible, test with a 1200W+ Platinum-rated power supply.

Reduce GPU Power Limit:

Try reducing the GPU's power limit (e.g., set it to 80% in MSI Afterburner) to see if this prevents the crashes.

System and Slot Testing:

Update the motherboard BIOS.

Try moving the graphics card to the second PCIe slot for testing.

These recurring nvlddmkm.sys + 0x116 errors on the RTX 4090 platform are almost always hardware-related and point away from purely software causes. Starting with the power delivery path will give you the highest chance of locating the problem.

Answer 3

All your minidump files, bar one do indicate the Nvidia drivers are causing your system to crash, the other does list ntfs.sys, but that is just a generic error, not the underlying cause of that crash, there is no other problem indicated on your system or with hardware, and the Nvidia drivers are a well known cause of system crashes.

You indicate you have tried different versions of the Nvidia drivers, did you first use the DDU utility to completely remove the current Nvidia drivers, and did you try a version of the Nvidia drivers around 572 (if available for your graphics card model).

Answer 4

The intermittent crashes you are experiencing with the VIDEO_TDR_FAILURE (0x116) bug check code can indeed suggest a hardware issue, particularly related to the GPU, PCIe, or power delivery. This bug check typically indicates that the graphics driver has timed out while waiting for the GPU to respond, which can be caused by various factors, including hardware instability.

Given that you have tested multiple NVIDIA driver versions and the crashes persist, it leans more towards a hardware issue rather than a software or driver bug. The fact that the crashes occur after long periods of stability and then suddenly lead to instability supports this theory.

Regarding known issues, while there are no specific references in the provided context about persistent 0x116 crashes across many driver versions, it is generally acknowledged that such behavior can be indicative of hardware problems rather than software bugs.

For additional testing on the Windows side, consider the following:

• Run hardware diagnostics: Use tools provided by your system manufacturer to check for any underlying hardware issues.
• Check power supply: Ensure that your power supply unit (PSU) is adequate for your GPU and system configuration, as power delivery issues can cause instability.
• Monitor temperatures: Use software to monitor GPU and CPU temperatures to ensure they are within safe operating limits.
• Test with a different GPU: If possible, testing with another GPU can help determine if the issue is specific to your current graphics card.

In summary, the pattern you described does suggest a hardware issue, and further diagnostics focusing on the GPU and power delivery would be prudent.

---

References:

• WDI information collection for bugs
• Advanced troubleshooting for stop code errors
• Bug checks (stop code errors)