Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

RuntimeError: Can't add devices across platforms to a single context. -33 (PI_ERROR_INVALID_DEVICE) #76

Open
DurianyDoriana opened this issue Oct 12, 2024 · 3 comments
Labels
bug Something isn't working

Comments

@DurianyDoriana
Copy link

Describe the bug

After installing AI-Playground-v1.21b I'm getting this error:

Backend initialization failed
The AI inference backend failed to initialize. Please try restarting the application. If the problem persists, you can check the Details for additional information about the error.

return _C._has_fp64_dtype(device)
       ^^^^^^^^^^^^^^^^^^^^^^^^^^

RuntimeError: Can't add devices across platforms to a single context. -33 (PI_ERROR_INVALID_DEVICE)

To Reproduce

Steps to reproduce the behavior:

  1. Install the recommended https://learn.microsoft.com/en-us/cpp/windows/latest-supported-vc-redist?view=msvc-170
  2. Install https://github.com/intel/AI-Playground/releases/download/v1.21beta/AI.Playground-v1.21b-Mobile-ICU-200v_LNL.exe
  3. Run AI-Playground
  4. See error

Expected behavior

Works just like AI-Playground-v1.01b-dGPU

Environment (please complete the following information):

  • OS: Windows11 23H2 22631.4317
  • GPU: SPARKLE ARC A770 TITAN OC
  • CPU: 13500
  • Version: AI-Playground-v1.21b
@Nuullll
Copy link
Contributor

Nuullll commented Oct 14, 2024

You can install https://github.com/intel/AI-Playground/releases/download/v1.01beta/AI.Playground-v1.01b-Desktop_dGPU.exe since your device is an Arc A770 discrete GPU, where IPEX 2.1.40+xpu is used.

Thanks for reporting this issue.

@cf418
Copy link

cf418 commented Oct 24, 2024

I have the same issue. A workaround for me was to set the environment variable set ONEAPI_DEVICE_SELECTOR=*:<id of dgpu> to prevent the iGPU from being seen.

@Nuullll
Copy link
Contributor

Nuullll commented Nov 8, 2024

I have the same issue. A workaround for me was to set the environment variable set ONEAPI_DEVICE_SELECTOR=*: to prevent the iGPU from being seen.

Thank you! The workaround has been implemented in 93602fe

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

3 participants