NVIDIA Jetson AGX Thor: Introduction & Getting Started

Follow Us On

Index

Next: Introduction/SoM Overview

NVIDIA Jetson AGX Thor

Introduction

The NVIDIA^® Jetson AGX Thor™ is the newest and most powerful member of the Jetson family, combining a 14-core Arm® Neoverse V3AE CPU, a Blackwell Architecture GPU, and 128 GB LPDDR5X memory. It represents a generational leap over Jetson Orin in both performance and features.

Thor launches with JetPack 7.0, a unified software stack integrating CUDA®, cuDNN, TensorRT™, Holoscan SDK, and more.

This page is the introduction to the Jetson Thor wiki. Here you will learn how the Jetson ecosystem is structured and how this wiki is organized to guide you from first evaluation to advanced development.

The Jetson Ecosystem

The NVIDIA Jetson platforms — including Thor — are built around three main components:

System on Module (SoM):

The SoM is the heart of Jetson, it contains the CPU, GPU, memory, accelerators (like PVA 3.0), and high-speed I/O. It's designed to be production-ready and embedded directly into final products. For Jetson Thor this SoM is the Jetson T5000 Module. [Learn more]

Carrier Board:

The carrier board exposes the SoM's capabilities by providing physical connectors, power delivery, and interfaces. The NVIDIA Developer Kit serves as the reference design. It is useful to evaluate the capabilities of the SoM in the prototyping phase of your project, while ecosystem partners (Auvidea, Connect Tech, etc.) offer carrier boards for different and specific requirements. [Learn more]

Board Support Package (BSP) & Software:

This includes device drivers, kernel modules, and JetPack 7.0 — the unified software environment providing access to Thor's hardware and NVIDIA AI SDKs. TODO Learn more

Together, these three elements — SoM + Carrier + BSP/Software — form the Jetson Thor platform.

Why Jetson AGX Thor?

Compared to its predecesor Jetson Orin, Thor provides:

More CPU power (14 Neoverse V3AE vs 12 Cortex-A78AE)
A stronger GPU (2560 CUDA cores, 96 Tensor cores, Transformer Engine)
Twice the memory (128 GB LPDDR5X vs 64 GB LPDDR5)
Higher bandwidth (273 GB/s vs 204 GB/s)
Much higher AI throughput (up to 2070 FP4 TFLOPs, ~7x more than Orin's 275 TOPS)
New features like Holoscan Sensor Bridge and Multi-Instance GPU (MIG)

This makes Thor ideal for complex computational and AI tasks like humanoid robots, autonomous machines, healthcare AI, generative AI at the edge, and real-time multimodal sensor fusion.

Wiki Structure & Navigation Index

🧹 jchavarria: These are potential sections. During development they might change — review this once the wiki is close to publishing and links are final. (remove this box when addressed)

This section is a main guide for the rest of the wiki. Each block below explains what the corresponding section covers and what you will find inside.

🖥️ Introduction

This section introduces the Jetson AGX Thor platform; it's the starting point for understanding how Thor fits into the Jetson lineup.

SoM Overview - Highlights the main characteristics of the T5000 SoM.
Carrier Boards - Describes Jetson Thor carrier boards.

⚙️ JetPack 7.0

Everything related to JetPack 7.0: installation, supported SDKs (CUDA, cuDNN, TensorRT, Holoscan, DeepStream, GStreamer), and tools for AI/robotics development.

Getting Started - Install JetPack, flash with the GUI installer, and explore included components.
Components - What's inside JetPack (CUDA, TensorRT, DeepStream, etc.) and example guides.
Compiling Source Code - Build kernel, device tree (DTB), and bootloader from source.
Flashing the Board from Cmdline - Flash the board manually via the command line.
Performance Tuning - Evaluate performance, tune power, and maximize efficiency.

🛠️ Development in the Board

Guidance for working on-device after JetPack is installed: access methods, package installs, and day-to-day development.

Getting into the Board - Use the serial console, SSH, and Ubuntu login.
Installing Packages - Install additional packages and development libraries.

🎥 Supported Cameras

Supported camera modules and drivers, setup instructions, and examples for capture/streaming.

IMX477+J20 Driver - Example sensor and driver integration.

🎮 Blackwell GPU

Overview of Thor's Blackwell GPU: CUDA/Tensor cores, Transformer Engine, and MIG (Multi-Instance GPU).

GPU Overview - Feature summary and architecture notes.
GPU Benchmarks - Performance results and Orin comparisons.

📡 GStreamer Pipelines

Examples and docs for capture, display, and encoding (H.264/H.265) on Jetson Thor.

Capture and Display - Run capture-to-display pipelines.
H264 Pipelines - Encode and stream with H.264.
H265 Pipelines - Encode and stream with H.265.

🔬 Holoscan

Real-time multimodal sensor fusion. Thor's Sensor Bridge enables Ethernet → GPU-direct streaming with low latency.

Holoscan Overview - SDK concepts and setup.
Holoscan Examples - Example applications on Thor.

👁️ PVA

Programmable Vision Accelerator (PVA 3.0): suitable workloads, performance, and CV pipeline integration.

PVA Overview - Architecture and capabilities.
PVA Usage - Practical acceleration examples.

🧩 RidgeRun Products

RidgeRun's optimized GStreamer plugins and utilities for Thor.

GstPTZR - Pan/Tilt/Zoom/Rotate plugin.
GstCUDA - GPU-accelerated processing framework.
GstColorTransfer - Color transformation utilities.
GstWebRTC - Peer-to-peer WebRTC streaming.
GstRtspSink - RTSP output element.
GstInterpipe - Cross-pipeline linking.
GstInference - Deep-learning inference integration.
GstVPI - NVIDIA VPI wrappers.
GStreamer Daemon - Headless pipeline control service.
GstShark - Profiling and performance analysis tools.

🧭 RidgeRun Reference Designs

Ready-to-use designs showcasing advanced Thor use cases.

Panoramic Stitching - Multi-camera 360° stitching reference.
WebRTC Streaming - Low-latency WebRTC streaming reference.

📖 Reference Documentation

Links to detailed documentation, developer references, and technical manuals for Thor.

Reference Documentation Main Page - Central entry point for specs and manuals.

✉️ Contact Us

Get support, services, and collaboration for Jetson AGX Thor projects. Reach RidgeRun for inquiries or assistance.