reComputer Jetson Platform Overview

Seeed NVIDIA Jetson Family

This chapter answers one core question: what exactly is Seeed's reComputer Jetson platform, and how does it relate to NVIDIA Jetson modules, carrier boards, full systems, and the software stack?

#Why This Chapter Matters

Before diving into JetPack, CUDA, TensorRT, and the AI projects in later chapters, it helps to separate three layers clearly:

  • Jetson module: the compute core provided by NVIDIA, including the CPU, GPU, NPU, memory, and other core processing resources.
  • carrier board: the platform provided by Seeed that exposes interfaces such as USB, CSI, Ethernet, CAN, and M.2.
  • full system: a deployable edge AI device built by combining the module and carrier board with thermal design, storage, enclosure, power, and BSP support.

#Chapter 2 Outline

ModuleTopicWhat You Will Learn
Module 2.1What Is reComputer JetsonUnderstand that reComputer is not a single model, but a product family built around Jetson SoMs
Module 2.2NVIDIA Jetson ModuleUnderstand the positioning differences between Orin Nano, Orin NX, AGX Orin, and Xavier NX
Module 2.3Seeed Jetson Compatible Carrier BoardCompare how different carrier boards emphasize different IO, power, and expansion capabilities
Module 2.4Jetson Full System SeriesUnderstand Seeed's product matrix through the Classic, Mini, Industrial, Super, and Robotics full-system families
Module 2.5Accessory SupportLearn the common accessories, interface types, and software ecosystem around reComputer Jetson

#Learning Goals

After this chapter, you should be able to answer:

  • Why TOPS is not the only metric that matters when choosing a platform
  • Why the same Jetson Orin NX can appear in multiple Seeed product families
  • Why some scenarios prioritize serial / CAN / dual Ethernet / 5G / GMSL2 more than raw compute alone
  • Which type of reComputer is a better fit if your goal is to learn JetPack and basic AI inference

#Suggested Reading Order

It is recommended to read this chapter in the following order:

  1. Start with 2.1 to build the overall mental model.
  2. Continue with 2.2 to understand the NVIDIA Jetson module lineup.
  3. Read 2.3 and 2.4 to distinguish carrier boards from full-system product families.
  4. Finish with 2.5 to connect interfaces, accessories, and the broader platform ecosystem.

#Official References