Retrieve The Oldest and Newest Supported CUDA Versions
See original GitHub issueFeature request
The STUMPY Python package depends on numba
and, when cuda
is available, it imports submodules that rely on cuda. It checks this by doing something like:
from numba import cuda
if cuda.is_available(): # Only import if CUDA is available
from stumpy import gpu_stump
However, numba
does not currently support CUDA 11.2 and if CUDA 11.2 is installed then the above if
condition still returns True
. It would be nice to have function that checks if the available CUDA version is compatible with numba
like:
if cuda.is_available() and cuda.version_is_supported():
from stumpy import gpu_stump
Or, as discussed with @stuartarchibald, maybe it makes more sense to add max/min supported versions enumerated and then make that part of the cuda.is_available()
check.
Issue Analytics
- State:
- Created 3 years ago
- Comments:7 (7 by maintainers)
Top Results From Across the Web
Older versions of Cuda - CUDA Setup and Installation
Is it possible to install older versions of Cuda anymore? I'd like 10.1 with a compatible version of gcc.
Read more >How to get the CUDA version? - Stack Overflow
You should find the CUDA Version highest CUDA version the installed driver supports on the top right corner of the comand's output.
Read more >How to find the NVIDIA cuda version - nixCraft
Open the terminal application on Linux or Unix. · Then type the nvcc --version command to view the version on screen: · To...
Read more >New CUDA application released - Asteroids@home
We are happy to present you our latest CUDA application v102.16! ... CUDA Version: 11.8 to get tasks for the new app (CUDA...
Read more >Which CUDA version should I install? Do they have back ...
Are the newer versions back-compatible? As per Nmath's suggestion, I went on to install CUDA from the Ubuntu repository as follows. $ sudo...
Read more >
Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free
Top Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
I would also agree with the approach of combining the version check with the driver check in
cuda.is_available()
to keep things simple for downstream libraries.Fantastic! Thanks, @gmarkall - please feel free to close this issue