-
Notifications
You must be signed in to change notification settings - Fork 2k
Cuda 11 doesn't work, Cuda 10 does #1439
Comments
I assume this is a ´Debian 10` system? There are several open issues with this already: However, I am still unable to reproduce the issue, so it is hard to debug. |
Sorry for the double issue- Github search failed me. It's a standard Debian 10, upgraded from Debian 9. I was trying to use kernel 5.9 previously, but downgraded after finding out the nvidia driver isn't out of experimental yet. I've also tried manual installation of the nvidia drivers (using .run) It looks like my issue is the same as #1424, and I have
|
Hi, I'm facing the same issue on my Debian 10 Testing with nvidia/cuda:11.x-base. Is there any news concerning this issue ? Thank you for your help |
I accidentally upgraded to kernel 5.9 again, so now I'm holding out until Debian 11 is released |
After the latest nvidia-docker and gpu driver updates I have my gpu working inside of the container. I still had to set |
@crobibero Note, removing the |
I understand, but keeping the
|
1. Issue or feature description
Unable to run cuda 11 container.
2. Steps to reproduce the issue
Host
3. Information to attach (optional if deemed irrelevant)
nvidia-container-cli -k -d /dev/tty info
uname -a
dmesg
N/A
nvidia-smi -a
docker version
dpkg -l '*nvidia*'
orrpm -qa '*nvidia*'
nvidia-container-cli -V
Nothing logged
Working:
docker run --rm --gpus all nvidia/cuda:10.0-base nvidia-smi
Not working:
docker run --rm --gpus all nvidia/cuda:11.0-base nvidia-smi
The text was updated successfully, but these errors were encountered: