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
cuda can't init [802] Call to cuInit results in UNKNOWN_CUDA_ERROR #7187
Comments
I wonder what CUDA libs it's picking up - can you give the output of |
😁😁😁 |
Thanks - sorry, I totally forgot that this doesn't output the location of the libcuda.so that it's using. Can you check what libcuda.so another working app is using, then check what libcuda.so Numba is picking up? If you're not sure how to do this, one way to do it would be to start the app under GDB, wait until it's started using CUDA, then break in it and give the command Could you try this and paste the output of |
(Created #7188 to remind me to do something about the location of libcuda.so not being shown) |
Sure.
|
changing A100 to V100 on same system:
mystics.... |
Thanks - I see the output for one application (Python / Numba?) but not for the other (the working application, I think) - can you give the output for the other as well please? |
Please close issue. |
Glad you solved it - many thanks for following up with an explanation of how to resolve the issue, too! |
Hello,
i'm running on azure instance: ND96asr_v4
ubuntu: 20.04 LTS
kernel: 5.8.0-1036-azure
numba: 0.53.1
no errors in dmesg
other apps works just fine.
exactly same disk been used with V100 and everything works smooth.
Any ideas what it can be?
The text was updated successfully, but these errors were encountered: