Issues with LiveKernelEvent 141 and 117. Persistant, regardless of what I do :(

Question

Issues with LiveKernelEvent 141 and 117. Persistant, regardless of what I do :(

Adam ‎ 40

Hello!

I've had recurring issues with my GPU/GPU driver. It started when I DDU'd in safe mode (up-to-date drivers crashed The Finals, downgrading drivers via AMD wouldn't allow me to overclock in MSI Afterburner). My PC would black screen and restart during gaming at seemingly random points, leaving behind hardware errors in reliability history (141 and 117).

What I've got:

AMD 7800XT, undervolted and overclocked (stable before issues, still black screen crashes after resetting to default settings in afterburner)
- Afterburner settings: 1050mv, +15 power limit, 2800MHz core clock, 2500 MHz memory clock, default Afterburner fan curve (I've kept my eye on the temps, and my GPU doesn't seem to thermal throttle or go much past 60-70c? Temperature numbers might differ, but it doesn't get too hot, just loud)
Corsair 750 watt PSU
1440p 165hz monitor
Windows 10 IoT LTSC

What I've tried:

Multiple follow-up DDUs, as well as updating to the newest AMD driver version
Windows reinstall while keeping files (wiped my appdata folder and saved game files, thanks Windows, but it's on me for not backing up my drive)
New Windows install via bootable USB, deleted old partitions, installed Windows, let it update stuff, then installed motherboard drivers and AMD drivers
- Transferred my Appdata stuff willy-nilly and corrupted Discord & Firefox
Another new Windows install, this time I installed the motherboard drivers and AMD drivers before Windows was even able to connect to wifi and mess anything up
- This time manually transferred important Appdata stuff, no corruptions, no old Windows files
I've reseated the GPU, as well as reconnected the daisy-chained extra power cables that connect to the GPU
- I haven't reconnected the cables to the PSU yet, nor tried connecting 2 different power cables to the GPU instead of the daisy-chained single cable

I couldn't find a minidump file after the crash today (first one since the most recent reinstall), but I've updated the minidump settings from automatic memory dump to small memory dump, so maybe it'll make one next time? I've tried to include everything that I think would be helpful, so please tell me what else is needed.

One thing that's new is:

Installation Failure: Windows failed to install the following update with error 0x80240009: SteelSeries ApS - HIDClass - 2.8.0.0.
Installation Failure: Windows failed to install the following update with error 0x80240017: Update for Windows Defender Antivirus antimalware platform - KB4052623 (Version 4.18.2001.10).
Installation Failure: Windows failed to install the following update with error 0x80240017: 2024-08 Update for Windows 10 Version 21H2 for x64-based Systems (KB5042320).

I was able to check for new Windows updates and install the Defender update fine, though.

And here's the 141 error:

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎13 3:37 AM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 141

Parameter 1: ffffbf8fc0e1c460

Parameter 2: fffff803c7ed7e40

Parameter 3: 0

Parameter 4: 2ef4

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105

I haven't had any installation failures before, so I'm confused. The USB I installed Windows on is pretty old, but it's the same one I used before I had these issues. My best guess was DDU corrupting my Windows somehow? But after all these weeks of issues, I'm lost and tired. I just hope it's not an actual hardware issue with my GPU. It's still under warranty, but sending it via mail to Gigabyte is anxiety-inducing.

Thanks for reading all this, and please help me lol

0 comments

Answer accepted by question author

0 additional answers

Your answer

Answer 1

Ian-Ng 14,140 Microsoft External Staff Moderator

Hi @Adam,

Welcome to Microsoft Q&A forum.

Thank you for providing the details regarding the persistent black screen crashes and the LiveKernelEvent 141 and 117 errors you have been encountering. These errors indicate a Timeout Detection and Recovery (TDR) failure, where the Windows operating system loses communication with the GPU driver.

Given that you have already performed multiple clean Windows installations and driver wipes (DDU), we should focus on power delivery and system configuration. Kindly follow these steps to isolate the cause:

1. Power delivery method

Please ensure the GPU is not powered by a single "daisy-chained" PCIe cable. For a high-draw card like the 7800XT, you need to use two separate PCIe power cables directly from the PSU. This prevents voltage drops during transient power spikes that frequently trigger driver timeouts.

2. Disable Multi-Plane Overlay (MPO)

MPO is a Windows feature that can cause conflict with AMD drivers, leading to black screens. I recommend disabling MPO via the Windows Registry or using a reputable MPO Disable tool.

3. BIOS and Software configuration

PCIe Link Speed: enter your Motherboard BIOS and manually set the PCIe slot speed to Gen 4 (or Gen 3 for testing) instead of "Auto."
Software Conflicts: Completely uninstall MSI Afterburner and reset AMD Adrenalin to factory default settings. Overlapping overclocking utilities can cause low-level instability.

If the issues persist after these adjustments, we may need to consider a hardware-level assessment of the GPU or PSU. Please let me know the results after testing these changes.

I appreciate your cooperation and look forward to your update.

If the answer is helpful, please click "Accept Answer" and kindly upvote it. If you have extra questions about this answer, please click "Comment".

Note: Please follow the steps in our documentation to enable e-mail notifications if you want to receive the related email notification for this thread.

Adam ‎ 40 Reputation points

2026-01-17T05:22:21.97+00:00

Hey, so none of that worked. I added an extra PCIe cable, none of them were loose either. I Disabled MPO with a registry file. Also, changed the PCIe link speed and PCIex16 link speed from auto to gen 4. I don't have the full AMD Adrenalin software, just the drivers and AMD Install Manager that it installs on its own, so there shouldn't be any overlapping overclocks. Is there any other things I can try? Or is my GPU messed up at the hardware level?

Some extra details I remembered:
After a game crash to desktop, my monitor's framerate would be halved. It's globally limited to 142fps via RivaTuner, and the only way to fix it is to launch another game for some reason. Win+Crtl+Shift+B doesn't fix it. I used to get a mix of full black screen crashes and game crashes, but now it's just crashing the game and halving the framerate. I don't have any minidumps since I haven't fully crashed still, didn't even have the folder in C:\Windows. Edit: Forgot to add that when the game crashes, my OBS also crashes. It's not unresponsive, but still isn't able to be closed without Task Manager.

I do have some new errors in Reliability History:

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎16 7:25 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 144

Parameter 1: 3003

Parameter 2: ffff9e0f2972f030

Parameter 3: 40010000

Parameter 4: 0

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎16 11:46 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 1a8

Parameter 1: 1

Parameter 2: 0

Parameter 3: 0

Parameter 4: 0

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎16 11:46 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 1b8

Parameter 1: 1

Parameter 2: 0

Parameter 3: 0

Parameter 4: 0

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎16 11:46 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 1b8

Parameter 1: 1

Parameter 2: 0

Parameter 3: 0

Parameter 4: 0

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105

Any ideas? Thanks for the help.
Ian-Ng 14,140 Reputation points Microsoft External Staff Moderator

2026-01-19T21:54:48.9+00:00
Hi @Adam,

Thank you for your update, the fact that your system no longer black-screens but instead crashes to a "halved framerate" suggests the driver is recovering, but the GPU is getting stuck in a low-power state.

Please try these specific steps to isolate the remaining issues:

Disable ULPS (Fix for Halved Framerate): In MSI Afterburner settings, find and check "Disable ULPS" (Ultra Low Power State), then restart. This prevents the GPU from "falling asleep" after a driver reset, which is what causes the performance drop.

Test without Hardware Encoding: Since OBS is hanging, your GPU's encoder might be the point of failure. Try gaming without OBS open. If you need to record, go to OBS settings and change the Encoder from AMD HW to x264 (CPU).

Swap the Display Cable: Error 144 is often a communication "handshake" failure. If you are using DisplayPort, try switching to a High-Speed HDMI cable to see if the crashes stop.

Underclock the VRAM: Install the full AMD Adrenalin Software (not just the driver). In the Performance tab, manually set the Max Memory Frequency to 2400 MHz. If this stops the 1a8/1b8 errors, the card's memory is unstable at factory speeds.

If these steps do not stop the crashes, it is highly likely that the GPU has a hardware defect (either the VRAM or the Video Encoder). Since it is under warranty, I recommend preparing for an RMA with Gigabyte, as the card is failing to maintain stability at stock settings.
Ian-Ng 14,140 Reputation points Microsoft External Staff Moderator

2026-01-20T22:43:08.74+00:00

Hi @Adam,

Hopy you are doing well.

May I know that have you had a chance to review my previous response?

I'm looking forward to your updates.
Adam ‎ 40 Reputation points

2026-01-20T23:25:19.88+00:00

Hi sorry,

I haven't had the time to do extensive testing lately. I've disabled ULPS, OBS was already on software encoding, and I have not yet tried changing the display cable. I've also found out that my C: drive is missing the boot partitions, and that they were installed on my D: drive instead (more info on my other question on my profile). It shouldn't be affecting my GPU right? I don't have plans to remove my D: drive, so I'm not gonna mess with that. I'm planning on monitoring the rail voltages to the gpu with HWMonitor, and look for anything messed up there. Thanks for all the help so far.
Ian-Ng 14,140 Reputation points Microsoft External Staff Moderator

2026-01-21T23:07:23.0833333+00:00

Hi @Adam,

Thank you for the update. It’s good to hear that ULPS is disabled and OBS is already using software encoding. To answer your specific questions and help with your next steps:

1. Does the Boot Partition issue affect the GPU?

No. Having your boot partitions on your D: drive is a common result of Windows installations where multiple drives were connected, but it will not cause GPU hardware errors (141, 144, etc.). It only means your PC won't boot if the D: drive is removed. You can safely leave this as is for now while we focus on the GPU.

2/ About Voltage Monitoring

Monitoring your rail voltages is a great idea to rule out the PSU.

Kindly watch the "GPU 12V Input Voltage." If you see this value dip significantly (below 11.4V) right before or during a crash, it points toward a power delivery issue from your PSU.

3/ Remaining variables

When you have a moment for testing, the display cable is still a prime suspect for that Error 144. Even if the cable looks fine, a bad handshake can cause the exact "device census" failures you're seeing. Swapping to a known-good HDMI cable for a session is a quick way to rule this out.

Take your time with the testing. The fact that the system is no longer full crashing is a positive sign that we are moving in the right direction.

Appreciate your cooperation and understanding.
Ian-Ng 14,140 Reputation points Microsoft External Staff Moderator

2026-01-22T22:33:10.3033333+00:00

Hi @Adam,

Have a good day.

May I know that have you had a chance to review my previous response?

Any updates would be greatly appreciated.
Adam ‎ 40 Reputation points

2026-01-22T23:06:40.33+00:00

Hii,

I just finished up testing (and got a crash) like 5 minutes ago. The voltage seems to have spiked slightly right before the crash, then continued staying around that level/dropping down to 12.024 or so a couple times, then back up. My pc's framerate didn't halve this time, most likely due to disabling ULPS in Afterburner settings. I did try using an HDMI, and didn't get a crash for a while, but that might just be because Elden Ring doesn't really crash me reliably. Outlast Trials does within maybe 15 minutes (randomly) on max settings with no upscaling enabled, so that's what I use to test. HMDI limits my monitor's hz to 144 down from 165, so that might play into it, too. I'll try the HDMI again with Outlast Trials this time, probably sometime later tonight. I even managed to get a clip of the crash, since OBS didn't crap the bed this time (due to disabling ULPS again?). https://streamable.com/6dg7ye

This was the unreal engine crash report:

Fatal error: [File:Unknown] [Line: 713] pResource->Map(Subresource, pReadRange, reinterpret_cast<void**>(&pData)) failed at D:/jenkins-steam-release-5-0/OPP/UnrealEngine4/Engine/Source/Runtime/D3D12RHI/Private\D3D12RHIPrivate.h:1213 with error DXGI_ERROR_DEVICE_REMOVED with Reason: DXGI_ERROR_DEVICE_HUNG

And another 141 in reliability history:

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎22 5:43 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 141

Parameter 1: ffff8b08191d21e0

Parameter 2: fffff80792dc7e40

Parameter 3: 0

Parameter 4: 131c

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105
Adam ‎ 40 Reputation points

2026-01-24T03:06:00.4066667+00:00

Hi again,

I just crashed again with an HDMI cable, so unless both are messed up, it's probably not the cable. Is there anything else to test?

The error:

Source

Windows

Summary

Hardware error

Date

‎2026-‎01-‎23 9:38 PM

Status

Not reported

Description

A problem with your hardware caused Windows to stop working correctly.

Problem signature

Problem Event Name: LiveKernelEvent

Code: 141

Parameter 1: ffff8b081b72c460

Parameter 2: fffff80792dc7e40

Parameter 3: 0

Parameter 4: 2ad4

OS version: 10_0_19044

Service Pack: 0_0

Product: 256_1

OS Version: 10.0.19044.2.0.0.256.191

Locale ID: 4105
Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.
Adam ‎ 40 Reputation points

2026-01-26T03:26:42.35+00:00

Hi, was that your comment that was deleted at the bottom? Could be mine, and I just forgot after re-posting the deleted stuff. Anyways, I'm still stumped about the crashes, so any help would be greatly appreciated lol.

Edit: Game crashed and my background became black. So that's new.
Ian-Ng 14,140 Reputation points Microsoft External Staff Moderator

2026-01-26T23:26:39.4466667+00:00
Hi @Adam,

The latest data from your crash specifically the DXGI_ERROR_DEVICE_HUNG message is very telling. This error means the GPU literally stopped responding to the engine's commands, forcing the game to close. Since this happened even on an HDMI cable and with ULPS disabled, we can rule out the display interface.

Here are the final conclusions and a few "last resort" steps:

1/ The Verdict on "Device Hung"

The fact that your voltage stayed stable around 12.024V (which is perfectly normal) suggests your PSU is doing its job. However, the DXGI_ERROR_DEVICE_HUNG often means the GPU core or VRAM is failing to maintain its factory-set clock speeds at the provided voltage.

2/ Final Testing Steps

Before deciding on an RMA, please try these two final adjustments to see if they "cure" the instability:

The "Power Limit" Test: Open your AMD Adrenalin software (Performance > Tuning) and reduce the Power Limit to -10%. If the crashes stop, the card’s silicon is unable to handle its own factory power spikes.

The "Underclock" Test: Manually lower the Max Frequency of the GPU core by 200MHz. If the card becomes stable at a lower speed, it confirms the hardware is defective as it cannot meet its advertised specifications.

3/ Regarding the Video Clip

The fact that OBS didn't crash this time is a direct result of disabling ULPS. This kept the driver "awake" enough to save the recording, even though the game engine itself died.

If the card still crashes after lowering the power limit and clocks, I strongly recommend starting the RMA process with Gigabyte. You have done more than enough due diligence to prove this is a hardware-level instability.

Thank you for your cooperation.
Adam ‎ 40 Reputation points

2026-01-27T00:01:39.1866667+00:00

Thank you for all the help this past however long lol. Even if the gpu is stable after underclocking, that still means there's something wrong hardware-wise, right? It was stable on this overclock/undervolt before though, did it just take some time to show factory defects? I really hope it's not some niche issue, and Gigabyte doesn't send it back after finding no issues. I'm kinda screwed then lol