NVIDIA-SMI couldn't find libnvidia-ml.so library in your system

Robotics & Edge Computing Jetson & Embedded Systems Jetson AGX Orin
,
1

Hello,

Here’s some information about my system: I have a Jetson AGX Orin 64GB with JetPack 6.2 installed on the board. I’m trying to run the nvcr.io/nvidia/l4t-jetpack:r35.2.1 container using the following command:

docker run -it --rm --net=host --runtime nvidia -e DISPLAY=$DISPLAY -v /tmp/.X11-unix/:/tmp/.X11-unix nvcr.io/nvidia/l4t-jetpack:r35.2.1

However, inside the Docker container, when I run nvidia-smi, I get the following error:

NVIDIA-SMI couldn't find libnvidia-ml.so library in your system. Please make sure that the NVIDIA Display Driver is properly installed and present in your system.
Please also try adding the directory that contains libnvidia-ml.so to your system PATH.

nvidia-smi on the host has following output

+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 540.4.0                Driver Version: 540.4.0      CUDA Version: 12.6     |
|-----------------------------------------+----------------------+----------------------+
4

Hi,

There are some driver updates between r35 and r36.
Please use the container with the r36 tag instead.

For example: nvcr.io/nvidia/l4t-jetpack:r36.4.0
NVIDIA L4T JetPack | NVIDIA NGC

Thanks.

6

But what can I use if I need a CUDA 11.x version and containers with the r36 tag have CUDA 12.x? Can I use the nvidia/cuda:11.7.1-cudnn8-devel-ubuntu22.04 image for Jetson? Or is there a way to install CUDA 11.x on the l4t-base?

7

Hi,

Usually the container that supports Jetson contains l4t or iGPU keywords.
Due to the driver compatibility, you will need to donwgrade to JetPack 5 to use CUDA 11.x.

Thanks.

8

So, does this mean that JetPack 6 and JetPack 5 are not compatible, and I must downgrade to JetPack 5 to use CUDA 11.x?

9

Hi,

To run an older CUDA on the newer GPU driver is not supported.
But running a new CUDA library on an old GPU driver should be okay.

So you will need to downgrade to JetPack 5 to run the CUDA 11.
Thanks.

10

Thanks for your response.

11

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.