**Intel Customer Support Team / Graphics Driver Engineers,**
I am writing to report a critical and reproducible bug affecting the **Intel Arc Graphics Service** when running **Microsoft Flight Simulator 2020 (MSFS 2020)** on a system with an **Intel Core Ultra 7 155H** processor (integrated Arc graphics).
After extensive debugging, I have identified that the issue is not merely a game crash, but a fundamental failure of the Intel graphics driver service itself, which perfectly explains the symptoms observed.
---
1. Problem Summary & Core Symptoms
- **Game:** Microsoft Flight Simulator 2020 (Steam version, running in **DX11 mode**).
- **System:** Intel Core Ultra 7 155H, Intel Arc Graphics.
- **Primary Symptom:** The game runs perfectly for **one complete flight session** (short or long haul). Immediately upon attempting to start a **second flight** (without restarting the PC), the game crashes to desktop.
- **The "Fix":** The only way to fly again is to **completely reinstall the Intel graphics driver** or perform a full system reboot. This temporarily restores functionality for another single flight, creating a repetitive "fly once, reinstall once" cycle.
- **In-Flight Hangs:** During the first (successful) flight, I frequently experience a **2-3 second complete screen freeze**. During this freeze, **audio continues to play normally** and the simulation logic appears to continue running. The freeze can be temporarily resolved by pressing `Win + D` (show desktop) and then switching back to the game. Upon returning, the game "fast-forwards" to catch up with the simulation time.
---
2. Key Evidence: Windows Event Logs
Analysis of the Windows Application Event Logs reveals the root cause, showing **two distinct errors occurring during the same flight session**:
Log 1: The Core Culprit - Intel Graphics Service Crash
This log proves that the Intel driver service itself is crashing due to an unhandled .NET exception.
- **Log Name:** Application
- **Source:** Application Error
- **Event ID:** 1000
- **Faulting Application Name:** `IntelGraphicsSoftware.Service.exe`
- **Faulting Module Name:** `KERNELBASE.dll`
- **Exception Code:** `0xe0434352` **(This is a .NET Runtime exception)**
- **Timestamp:** 2026/2/18 11:16:48
Log 2: The Resulting Game Crash
This log shows the inevitable game crash that occurs when it tries to communicate with the now-unstable or crashed driver service.
- **Log Name:** Application
- **Source:** Application Error
- **Event ID:** 1000
- **Faulting Application Name:** `FlightSimulator.exe`
- **Faulting Application Name:** `FlightSimulator.exe`
- **Exception Code:** `0xc0000005` **(Access Violation - memory access error)**
- **Timestamp:** 2026/2/18 13:09:45
---
3. Technical Analysis: Explaining the "Fly Once" Pattern
Based on the evidence, I have reconstructed the failure sequence:
1. **The Trigger:** During the high-load rendering of the first flight, the game issues a graphics command that exposes a bug within the `IntelGraphicsSoftware.Service.exe`. The service enters an unhandled .NET exception state (`0xe0434352`) and crashes. The crash is logged at 11:16:48.
2. **The "Zombie" State:** Even though the service crashes, the Windows Display Driver Model (WDDM) may attempt to restart it. However, the restarted service is now in an unstable, "dirty" state. Critical GPU resources and memory may be improperly released or corrupted.
3. **The Freeze & Recovery:** The game's render thread hangs while waiting for the dead driver, causing the screen freeze. Because the audio and logic threads are separate, they continue to run. The `Win + D` trick forces a mode switch in the Desktop Window Manager (DWM), which can sometimes unblock a stuck GPU pipeline, allowing the game to resume rendering at an accelerated pace to catch up with the simulation logic.
4. **The Inevitable Second Flight Crash:** When a **new flight** is loaded, the game attempts to establish fresh, complex rendering pipelines with the driver. It communicates with the already corrupted driver service, which leads to an illegal memory access (`0xc0000005`), crashing `FlightSimulator.exe`. This is logged at 13:09:45.
5. **The "Reinstall" Fix:** Reinstalling the driver completely kills the corrupted service and its residual memory state, replacing it with a fresh, clean instance. This resets the cycle.
---
4. Request to Intel Development Team
I respectfully request that your engineering team investigate the following:
1. **.NET Exception in `IntelGraphicsSoftware.Service.exe`:** Please analyze the code paths within the driver service that are invoked during MSFS 2020 rendering (specifically in DX11 mode). The unhandled `0xe0434352` exception indicates a flaw in the service's managed code that must be caught and handled gracefully.
2. **TDR and Service Recovery:** Investigate why the service fails to recover cleanly after a crash or a Timeout Detection and Recovery (TDR) event. The current behavior leaves the GPU in an unstable state, requiring a full driver reinstall.
3. **Driver Version:** This issue has been observed on the latest driver, **32.0.101.8509 WHQL**. It was also present, though manifested differently, in version **32.0.101.8247**. The most stable driver tested so far is **31.0.101.5234 WHQL**, which does not exhibit this service crash.
This bug severely impacts the usability of the Intel Arc graphics on my system for professional flight simulation. Providing a fix in a future driver update would be greatly appreciated.
Please let me know if you require any further logs or system information.
Thank you for your time and for developing great hardware.
Sincerely,
蓦念Monnie
链接已复制
Hello 蓦念Monnie,
Thank you for posting on Intel Community Forum.
I understand that you're experiencing issues with Microsoft Flight Simulator 2020, and I appreciate you taking the time to share all the details you have so far. Your information is very helpful as I continue to troubleshoot this.
I also noticed that you reached out to us through an internal channel. To prevent any confusion or duplication of efforts, could you please let me know which communication channel you prefer to continue using for support?
I look forward to your response.
Best regards
Jed G.
Intel Customer Support Technician
Hello 蓦念Monnie,
I understand your preference to use internal support channels, so I'll proceed with closing this thread. For any future assistance, please submit a new thread as this one will no longer be active.
Sincerely,
JeanetteC.
Intel Customer Support Technician
