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.

Sign up for GitHub

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

Jump to bottom

[cuda] CUDA warp level intrinsics failed on legacy GPU #4935

Closed
0xzhang opened this issue May 9, 2022 · 5 comments
Closed

[cuda] CUDA warp level intrinsics failed on legacy GPU #4935

0xzhang opened this issue May 9, 2022 · 5 comments
Assignees
@qiao-bo
Labels
cuda CUDA backend potential bug Something that looks like a bug but not yet confirmed

Comments

@0xzhang
Copy link
Collaborator

0xzhang commented May 9, 2022 •
edited
Loading

Related PR is #4927. I shared the full error info on Gist.
It seems test_simt.py not testable on my PC.

  • GeForce GTX 1650
  • CUDA 11.4

I use command below.

python tests/run_tests.py -t1 -k test_simt

image

Warp functions have requirements to compute capability of device. Maybe my card doesn't have enough compute capability according the list.

If taichi could capture this problem in advance would be even better.
@0xzhang 0xzhang added the potential bug Something that looks like a bug but not yet confirmed label May 9, 2022
@taichi-ci-bot taichi-ci-bot moved this to Untriaged in Taichi Lang May 9, 2022
@0xzhang 0xzhang mentioned this issue May 9, 2022
Merged
@0xzhang 0xzhang added the cuda CUDA backend label May 9, 2022
@qiao-bo qiao-bo moved this from Untriaged to Todo in Taichi Lang May 13, 2022
@neozhaoliang neozhaoliang moved this from Todo to In Progress in Taichi Lang May 13, 2022
@ailzhang
Copy link
Contributor

Hey @0xzhang , thanks for reporting the issue! FYI @qiao-bo is OOO this week, so please expect a delay in his reply ;) Thanks a lot!

@0xzhang
Copy link
Collaborator Author

0xzhang commented May 20, 2022

Hey @0xzhang , thanks for reporting the issue! FYI @qiao-bo is OOO this week, so please expect a delay in his reply ;) Thanks a lot!

Thank you for telling me. You are welcome! 😄

@qiao-bo
Copy link
Contributor

qiao-bo commented May 27, 2022

Hi, sorry for my late reply. We should indeed detect this in advance by inspecting the GPU CC, we will fix this.

@qiao-bo qiao-bo mentioned this issue Jun 1, 2022
Merged
@qiao-bo
Copy link
Contributor

qiao-bo commented Jun 1, 2022

This should be fixed by 5077. Feel free to reopen if you encounter further problem.

@qiao-bo qiao-bo closed this as completed Jun 1, 2022
Repository owner moved this from In Progress to Done in Taichi Lang Jun 1, 2022
@0xzhang
Copy link
Collaborator Author

0xzhang commented Jun 5, 2022

This should be fixed by 5077. Feel free to reopen if you encounter further problem.

Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@qiao-bo qiao-bo

Labels
cuda CUDA backend potential bug Something that looks like a bug but not yet confirmed
Projects
Taichi Lang
Status: Done
Milestone
No milestone
Development

No branches or pull requests
3 participants
@ailzhang @qiao-bo @0xzhang