DXGI_ERROR_DEVICE_REMOVED or DXGI_ERROR_DEVICE_HUNG

These crashes were also plaguing me, on a 9060xt as well. Tried lots of troubleshooting and so called fixes but nothing worked. 

The game would crash after 10 mins into a round, after an hour, could go 3 -4 crashes in a row within minutes or hours without a crash. Certain maps it seemed to happen on more like New Sobek city and Iberian Offensive but it did happen on  other maps just less often so I took those out of rotation. 

I decided to test a 5060ti 16gb card and played with 0 crashes. Out of interest I reinstalled the 9060xt on driver 26.2.2 and have not crashed in 40-50 hours of gameplay now. I did not change anything else, using the exact same settings in Adrenalin, in game and in bios. 

Alright guys... free week is over and i played almost 13 hours without any DirectX crashes. I haven't changed anything on my system since the last crash where i refunded the game, with the exception that i now have -30/-150mhz on CO on my Ryzen 7600 (it was previously on -20), everything else is stock. This is tempting me to rebuy the game, HOWEVER it's clearly not fixed since other people are experiencing it.

For all of you still getting Directx crashes, on which maps do they usually occur? Any of the new season 2 maps? How frequent are they still?

It would be really nice for "old" users who experienced this months ago to come forward and tell if they are still getting crashes (if they're still playing LOL).

PS: I have a 9060 XT with the 26.2.2 adrenalin driver.

NOPE, DX11 put a BBC in my mouth. Somebody removed my GPU again, so for me DX11 mode no fix.

This appears to have resolved the issue for me, although I can't be fully sure until I've gone at least 10 hours without a crash. Thanks for the tip. 

Everyone should try this to see if it works. I haven't noticed any degraded performance or missing features after switching to DX11, although I've had all AMD features off for a while so I don't know if those still work.

The only reason DICE updated the SDK files was most likely to allow RX 9000 series GPUs to run the FSR 4.1 upscaler (instead of the old 4.0.2 one). In my opinion, I doubt they even noticed that issue which I mentioned.
Furthermore, going back to what you said: when I’m playing on the Manhattan Bridge map, I experience some awful stutters that occur at random times. So yeah, the maps themselves could be the cause (I have no doubt about that). As you can tell, the new map ‘Hagental Base’ is much more polished than the nine maps from Season 1. So a big factor could definitely be the maps themselves.

**EA ID / Steam ID:** r0ckstar87

**OS Version:** Windows 11 (Build 26100.1)

**GPU:** AMD Radeon RX 7900 XTX

**GPU Driver:** Tested on both 26.2.2 (25.30.27.02) and 26.3.1 (25.30.33.01) — crash reproduces on both

**Date/Time the issue occurred:** Multiple sessions: Mar 11 21:58, Mar 11 22:27, Mar 20 22:12, Mar 20 23:12

**Brief summary of your bug:** Reproducible DXGI_ERROR_DEVICE_HUNG crash occurring across multiple sessions, driver versions, and game versions. I have captured four crash dumps which all share an identical failure hash ({577a89c1-cd47-d3d4-2233-26f4a7549d0f}), indicating a deterministic bug in the Frostbite DX12 render pipeline rather than an environmental or hardware issue. GPU telemetry logs (GPU-Z, 1-second resolution) confirm the card is healthy at the time of each crash — one crash occurred at only 52°C hotspot, 66°C memory, and 122W power draw (~25% of TDP).

**Type of issue:** Crash to desktop with DirectX error dialog

**Mode where it occurred & Map:** Multiplayer — occurs across multiple maps

**Any other errors:** DirectX error dialog on every crash:

"DirectX function 'm_submitFencePlat->GetCompletedValue()' failed with DXGI_ERROR_DEVICE_REMOVED. DeviceRemovedReason: DXGI_ERROR_DEVICE_HUNG ('The GPU will not respond to more commands')."

---

**Additional technical detail for the dev team:**

All four minidumps produce an identical call stack terminating in Frostbite's DX12 submission thread:

• Exception code: 0x887a0005 (DXGI_ERROR_DEVICE_HUNG)

• Failure bucket: XAML_887a0005_bf6.exe!Unknown

• Failure hash: {577a89c1-cd47-d3d4-2233-26f4a7549d0f} — identical across all crashes

• Faulting module: Engine_Render_Core2_PlatformPcDx12_retail

• Key stack frames: bf6!ffxFsr2ResourceIsNull → rc2DeviceInit → rc2GetSystemServiceProvider

Note: FSR/upscaling was disabled for all sessions. The ffxFsr2ResourceIsNull symbol appearing in the call stack with FSR off may point to an unguarded FSR 2 code path executing regardless of the in-game setting.

**Confirmed NOT causal:**

• GPU temperature (crashes at 38°C GPU / 52°C hotspot)

• GPU load (crashes at 45% load, 63W)

• Fan speed / thermal throttling

• FSR / upscaling (disabled)

• Driver version (reproduces on 26.2.2 and 26.3.1)

• Game version (reproduces on 1.0.409.59773 and 1.0.414.19397)

**Partial workaround:** Enabling V-Sync reduced crash frequency in one session but did not eliminate the issue.

I have three .mdmp files and full GPU-Z sensor logs for all four sessions available if useful for the team.

Played yesterday and today for about 5ish hours with FSR on balanced mode and didn't get a single error. AMD anti-lag was also on. However, most of my playtime have been on the newer maps available on the free week. The only "OG" i got was the golf course one (i think it's the only one available). 

Dunno if the map pool has anything to do with this error but it's something to consider.

Anyway, i think it's too soon to say it has been fixed. A patch note would have been welcome.

What do you think?

After a lot of testing, I was finally able to narrow down the issue on my system, although I still don’t fully understand the root cause.

My System:

• CPU: AMD Ryzen 9 5900X
• GPU: PowerColor RedDevil RX 7900 XTX
• RAM: 32GB G.Sklill DDR4-3600
• Motherboard: ASUS ROG Strix X570-E Gaming

On my system, the game runs completely fine when I restrict the CPU affinity of EAAntiCheat.GameServiceLauncher.exe to CCD0 (cores 0–11) using Process Lasso. In this configuration, I can join matches and play without any crashes.

However, when I use CCD1 (cores 12–23) or both CCDs (cores 0–23), the game crashes again as usual while loading a map (around ~30% loading progress), without any error message.

This behavior is 100% reproducible.

So at this point it seems like:

• The game is stable on CCD0
• But becomes unstable when running on CCD1 or both CCDs

I don’t know whether this is a scheduling issue, an engine problem, or something related to anti-cheat, but it clearly seems to be linked to specific CPU cores / CCD usage on my system.

That being said, I can’t completely rule out a hardware-related issue on my side. It’s possible that something is wrong with the CPU (e.g. communication between CCDs), although synthetic stress tests (Prime95) did not show any instability when testing CCD1.

Recent updates fixed my problem after 3 days and still no crashes! 

What did I do?

Updated AMD 5900X chipset driver

Updated Win11 OS

Updated 9060XT 16Gb drivers

Updated BF6 to the new map and fixes.

From the above I ignore what did the trick or if it was a combo of things.

I had stopped played BF6 altogether on my AMD (no issues with NVidia) since my last message (month or more ago)

Hopefully you guys get lucky, and please report your successes, people tend to report the negatives and never mention positives.

Man i'm back playing since it's free now (refunded the game after 20 hours because of constant crashes). Played for over 2 hours yesterday with no crashes, but i had some settings turned to low-medium and FSR off. I will try again today and give you a feedback if everything works fine.

Thanks BTW!