My cuda is 12.2, how should I set up dGPU

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, 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)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

my graphics card is A40, the cuda version is 12.2, and the driver version is 535.54.03. now I want to run deepstream6.1.1 on this machine. I refer to the official installation tutorial (Quickstart Guide — DeepStream 6.1.1 Release documentation), and replace the second installation of CUDA Toolkit 11.7.1 with 12.2 (CUDA Toolkit 12.2 Update 1 Downloads | NVIDIA Developer), but in the third step, when installing TensorRT, I don’t know how to execute the command.

the cuda of the machine may not be changed, it can only be 12.2, and in my case, the software version is relatively high, can I run deepstream6.1.1?

please help me!

Is docker an option for your case?
DeepStream | NVIDIA NGC

thank you for your reply, docker may not be suitable for me:
1.the volume of the docker packaged program is too large
2. it is easy to see the source code

the background is like this:
i am using pyinstaller to package a streaming program developed based on deepstream. I perform the packaging operation on my 3090 test machine, and it is normal to run the packaged binary file, but uploading the packaged binary file to the A40 server fails. error: ‘streammux is None’, I don’t know why this happens, logically speaking, pyinstaller should package all the dependencies required by the program into the execution file and directly transplant it to a different machine to run. But at present, an error is reported after transplantation. I feel that some system dependent environments cannot be packaged into the file, so I want to install the deepstream6.1.1 operating environment on the A40

My understanding on the package from pyinstaller includes all necessary files for python, but please be noted that the host need satisfy DeepStream 6.1 prerequisites as described in Quickstart Guide — DeepStream 6.1 Release documentation (, this is not satisfied on your A40 environment.

do you mean that the graphics card platform is not satisfied? or is the graphics card driver and cuda version on my A40 not satisfied?

The GPU is OK, but the CUDA and Driver are NOK for DS6.1.

ok,i got it,thanks again for your help

i would like to confirm with you again, the A40 server ubuntu is 20.04.6, the driver version is 535.54.03, and the cuda version is 12.2. Is this configuration compatible to run deepstream6.1.1?
our colleagues in the procurement department said that when purchasing the Nvidia A40 server, the software version of the machine they delivered was as described above, and the people from Nvidia said it was compatible, so now I am confused, or is it not compatible with 6.1.1? but it is compatible 6.2?

If you want to install native DeepStream 6.1.1, please follow the dependency here: Quickstart Guide — DeepStream 6.1 Release documentation.

But it should be fine if you run deepstream docker on your existing environment.

ok, I understand now, thanks

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