oneAPI Causes Restart or Shutdown Failure After Running Ollama-Based Models

[yipingw]

I’m not sure whether this post is in the correct location, but I’m confident that this issue is caused by oneAPI.

I have an HP Omen Max 2025 laptop equipped with both an Intel integrated GPU and an NVIDIA discrete GPU (RTX 5080).

Recently, I downloaded and installed the full Intel integrated graphics driver package from Intel’s official website. During the installation, several additional components were installed as well, including oneAPI. After installing oneAPI, I ran a model based on Ollama via the PowerShell command line (in my case, gpt-oss:20b). After that, when I attempted to restart or shut down the computer, the operation only completed after a very long delay. Once I re-entered the Windows 11 operating system, a StackDump error log file appeared in the root directory of the C drive. This log indicates that a serious error occurred during the previous restart or shutdown. The detailed contents can be found in the attached file.

I then began troubleshooting to identify the cause. I uninstalled potential related software one by one, and after each uninstallation, I repeated the same steps: running a model via Ollama and then restarting the system. After numerous attempts, I found that once oneAPI was uninstalled, the issue disappeared.

I hope the information I’ve provided is helpful in resolving this issue. Please let me know if any further assistance or details are needed.