Simulation thread deadlock — game hangs a few minutes into matches (consistent stack)

Build Number: 16.2.10884.0 (also observed on 15.4.8719.0 — persists across updates)

Platform: Steam

Operating System: Windows 11 (25H2, build 26200)

Issue:
The game hard-hangs during matches. In single-player the units freeze while the mouse cursor and menu still respond; the window then goes “Not Responding” and Windows closes it after ~1 minute. In ranked multiplayer it drops to a white screen. No error code is shown; Windows logs it as APPLICATION_HANG_BusyHang for RelicCardinal.exe.

I captured full memory dumps during the hang with ProcDump and analyzed them in WinDbg. The signature is identical across 8 separate hangs:

  • “Simulation Thread” is blocked in ntdll!NtWaitForSingleObjectKERNELBASE!WaitForSingleObjectExRelicCardinal.exe at a consistent address — waiting on a sync object that is never signaled (a lost-wakeup race).
  • The Main and Render threads are then blocked downstream on SleepConditionVariableSRW, waiting for the simulation.
  • In ranked 1v1 the Main thread blocks at a different offset (possibly the networking/lockstep path).

Trigger Event:
No consistent trigger. The hang occurs at unpredictable times and is not tied to any specific action, Age-up, match length, or load. It has happened right at the start of a ranked 1v1, ~7 minutes into a co-op/PvE match, and at random points elsewhere — while some long (1+ hour) 4v4 matches complete with no hang at all. Started around December 2025 and continues through every update. The randomness is consistent with a timing-dependent race condition rather than a specific event.

Repro Steps:
Not reliably reproducible on demand — it is intermittent/non-deterministic. It occurs across all modes I play:

  • Ranked 1v1 multiplayer (sometimes within the first minute)
  • Skirmish / co-op vs AI (e.g., ~7 minutes in)
  • Large 4v4 matches (though many long ones finish fine)

Frequency varies session to session; there is no reliable set of steps that forces it. It has occurred consistently since ~December 2025 across multiple game builds.

  • Ruled out on my end (extensive testing): GPU driver version (multiple), Steam/NVIDIA overlays, RAM/EXPO (tested at JEDEC 4800), shader cache clear, Virtualization-Based Security / Hyper-V / Memory Integrity (disabled — no change), thermals/power. Other DX12 games run fine.

Specs: Ryzen 5 7600, RTX 4070 (driver 596.36), 32 GB DDR5, Win11 Pro 25H2, DirectX 12 / WDDM 3.2.

.dmp files

I installed Linux as a second system and have been playing for a week now. I’ve played 15 games and haven’t crashed once. It looks like some kind of Windows-related issue, but other games work fine on Windows.

After a week of playing, I got a freeze again. I asked the neural network to see what was happening. I have an internet connection, but one of the streams synchronizing settings with the Microsoft server froze. It seems I’m not the only one. This guy has the same problem. How can I solve this?

If the developers are watching this, the last time I was almost instantly kicked out was when I changed the button settings in the middle of the game (set the mouse wheel to select all cavalry), clicked save, and after 10 seconds the game froze and close.

I’m in Ukraine, and the ping to Microsoft servers is quite high. I’m offered the game to not survive even the slightest network connection while maintaining the settings.

Today I got two new freezes, the first one just in the middle of watching a replay, and the second one in a new campaign. Interestingly, in the campaign I can click on units but they don’t move!! I even recorded a video of it.