I have a jetson nano developer kit and I intend to run deepstream for production on it. Is it recommended to use docker or directly use tar.gz and use deepstream . What do you recommend ? I thought about this issue as resources are limited on nano unlike dGpu servers
Hi hrsht.sarma,
To run DeepSteam SDK directly on Nano is more straight forward than use docker at limited system resource and less performance impact.
1 Like
Does it means that using a deepstream container, Jetson Nano will have lower performance compared to a direct deepstream without docker?
There is any documentation or study which demonstrate these performance differences with and without docker?