Basic info of NVIDIA hardware suitable to deploy Deep learning algorithms

I’m required to use NVIDIA hardware for object detection related applications utilizing Deep Learning algorithms. As a result, I was looking for NVIDIA hardware that needed to be installed specifically for these applications to run. I investigated NVIDIA hardware GPU-based platforms (such as tegra K1, Jeston TK1, Jeston TX1, Jeston TX2, Jeston Nano, Jeston Xavier NX, Jeston AGX Xavier) for this purpose, but I need to clarify the few points.

  1. Do these boards contain built-in Operating Systems (such as Linux, mac, etc) or they require the user to install the operating systems first?

  2. Can i install following tools on the above mentioned NVIDIA hardware?
    i. CUDA Toolkit (>10.0)
    ii. cuDNN (> 7.5.0)
    iii. Tensorflow (> 2.0)
    iv. pytorch (latest versions)

  3. Do these boards contain ports (such as Ethernet) or any other ports to be compatible to receive the audio/video/text data packets directly? Data packets need to be received both from wired and wireless mediums.

Also, apart from above mentioned NVIDIA hardware boards, can you please recommend any recent and more suitable board/hardware?

I will be extremely appreciative of your kind response in this regard. Your thoughtful and affirmative response will be greatly appreciated.

Kind Regards,

Please refer to JetPack SDK | NVIDIA Developer

Please refer to below Jetson devkit and module info:
Jetson Developer Kits | NVIDIA Developer
Jetson Modules | NVIDIA Developer

Many thanks for your generous response. I got the necessary information.

Kind Regards,

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