Product Name & Persistence-M show ERR! in nvidia-smi with RTX A5000 on Deepstream 6.0

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
GPU : RTX A5000
• DeepStream Version
Deepstream 6.0
• JetPack Version (valid for Jetson only)
No
• TensorRT Version
TensoRT 8.0.1
• NVIDIA GPU Driver Version (valid for GPU only)

  • NVIDIA driver 470.63.01
    • Issue Type( questions, new requirements, bugs)
    Bugs
    • How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
    To reproduce the issue, the following command is provied
  • Driver installation command
dkms build-essential linux-headers-generic -y
./NVIDIA-Linux-x86_64-470.63.01.run --no-x-check --silent --dkms
  • CUDA installation command
cuda_11.4.1_470.57.02_linux.run --silent --toolkit --samples
echo 'export PATH=/usr/local/cuda-11.4/bin:$PATH' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=/usr/local/cuda-11.4/lib64:$LD_LIBRARY_PATH' >> ~/.bashrc
source ~/.bashrc
  • CuDNN installation command
tar -xzvf cudnn-11.4-linux-x64-v8.2.4.15.tgz
cp -P cuda/include/cudnn.h /usr/local/cuda-11.4/include
cp -P cuda/lib64/libcudnn* /usr/local/cuda-11.4/lib64/
chmod a+r /usr/local/cuda-11.4/lib64/libcudnn*

After that, I type ‘nvidia-smi’ to check the installation result and the product name space show ERR! and Persistence-M show Off as attached image
Screenshot 2024-04-05 at 12.49.22 PM
I have checked the deviceQuery and the follow error be shown as below:
Screenshot 2024-04-05 at 12.55.03 PM

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

  • If I upgrade my driver version to 5xx, the problem can be solve. However, I want to run Deepstream 6.0 on RTX A5000 if I use higher driver version I cannot install Deepstream 6.0 on it and as I checked the nvidia driver 470.63.02 also support RTX A5000 seri from download driver link
  • I hope you can help me to solve the problem while still use nvidia driver 470.63.02
    Thank you for your time I and hope to hear from you soon.

The 470.63.01 driver already installed the CUDA11.4, why do you need to install another CUDA manually?

Thank you for your reply. Since I follow installation structure of Deepstream 6.0.1 as it mentioned, so I download driver and CUDA separately. I have process this one on GTX 1080Ti , RTX 3080 and Tesla T4. They have no issue as RTX A5000


Besides that, I download and process it on Driver 470.63.01only but it still show the same error as I mentioned in the topic

I can reproduce the failure with RTXA5000. Move to the CUDA forum.

Hi Fiona.Chen, thank you for your information. Do you have any solution to cover the issue from RTX A5000. So I can continue use it with DS 6.0. Thank for your time and I hope to hear feedback from you soon.

Hi @Fiona.Chen , I have checked with our company again about the RTX seri, I think I have mistake between RTX A5500 version and RTX A5000 version. Since our product is RTX A5500 version and it not support for driver 470.63.02. Thank you for your support and I so sorry for this misunderstanding

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