[Fixed?] DXGI_ERROR_DEVICE_HUNG

Recently I came across this infamous editor crash, while placing meshes.

DXGI_ERROR_DEVICE_REMOVED with Reason: DXGI_ERROR_DEVICE_HUNG

Then I did the following differently

  • Saved more often
  • Disabled Allow Cinematic Control

A few hours later there are much more items in the viewport and no crash occurred. So perhaps this is useful. Otherwise you can try the following

Today I was again placing meshes at the same level spot, and it again crashed. Then I again saved a few times during mesh placement and no more crash occurred.

This suggest that the crash is related to the memory, while placing new meshes and that it can be overcome by saving.

Bit of a necro, but as much trouble as I went through to fix this error, figured I’d share.
In my case, it was one particular game where I had the error - everything else always ran perfect. I tried all the suggestions that I ran into - no help - and “knew” that my video card wasn’t overheating - temps maxxing in the 70s with hotspot in the low 90s.
Did some further testing and found that the delta for the gpu hotspot was higher than I like - at times, over 25 degrees - now, I know some folks say hotspot doesn’t matter and that it’s calculated - but in my experience when it starts moving up it’s a sign that your thermal paste / fans whatnot isn’t doing the job.
So, replaced the thermal compound with PTM - now, my temps doing the EXACT same application and settings went from 75 to 66, and hotspot from 94 to 79 - and ZERO crashes in 5 hours of testing (would crash at least once an hour before, if not more).
So, in my case, even though my temps were “fine” - they WERE still the problem.

1 Like

This is a bit of a throwback, but considering how much effort I put into solving this issue, I thought it’d be worth sharing.
For me, the problem only showed up with one specific game — everything else ran flawlessly. I went through all the usual troubleshooting advice without any success. I was confident my GPU wasn’t overheating, with core temps peaking in the 70s and hotspot temps staying in the low 90s.

Digging deeper, I noticed the hotspot delta was bigger than I liked — sometimes over 25°C. Now, I know there’s debate over how much hotspot temps actually matter since they’re calculated, but from my experience, when the delta grows too wide, it’s often a red flag that your thermal paste or cooling setup isn’t performing properly.

I ended up replacing the thermal paste with PTM, and the difference was night and day. Running the exact same game with the same settings, my core temps dropped from 75°C to 66°C, and hotspot temps dropped from 94°C to 79°C. Best part — zero crashes after 5 hours of play, whereas before I’d crash at least once an hour.

In short, even though my temps looked “okay” on paper, cooling issues were still causing the crashes. And once I got things sorted, I finally had smooth gameplay — just like streaming music on Spotify mod without a single skip.

1 Like