NVIDIA Jetson Platform

Find the Right Jetson for Your Edge AI Build

Compare Jetson platforms by product tier, compute performance, and best-fit edge AI applications.

Jetson AGX Orin
Selected platform

Jetson AGX Orin

Flagship Up to 275 TOPS Compute Power 100%
Contact Us

Specifications

Jetson AGX Orin
Module
NVIDIA Jetson AGX Orin 32GB
AI Performance
200 TOPS
GPU
1792-core NVIDIA Ampere GPU with 56 Tensor Cores
GPU Max Frequency
930 MHz
CPU
8-core Arm Cortex-A78AE v8.2 64-bit
CPU Max Frequency
2.2 GHz
DL Accelerator
2x NVDLA v2
Vision Accelerator
1x PVA v2
Memory
32GB 256-bit LPDDR5
Memory Bandwidth
204.8 GB/s
Storage
64GB eMMC 5.1
Video Encode
1x 4K60 H.265, 3x 4K30 H.265, 7x 1080p60 H.265
Video Decode
1x 8K30 H.265, 3x 4K60 H.265, 7x 4K30 H.265
Camera
Up to 6 cameras, 16 lanes MIPI CSI-2, D-PHY 2.1 / C-PHY 2.0
PCIe
Up to 2 x8 + 1 x4 + 2 x1, PCIe Gen4
USB
3x USB 3.2 Gen2, 4x USB 2.0
Networking
1x GbE
Display
1x 8K60 multi-mode DP 1.4a / eDP 1.4a / HDMI 2.1
Power
15W - 40W
Mechanical
100 mm x 87 mm, 699-pin Molex Mirror Mezz

Best-fit applications

Jetson AGX Orin
Advanced Autonomous Robotics
Large AI Model Inference
High-Performance Vision Systems
Multi-Sensor Edge Computing
Flagship Industrial AI Platforms

Why Jetson Matters

NVIDIA Jetson stands out because it combines GPU-accelerated AI compute, advanced camera and vision processing, robotics-ready software, and a mature deployment ecosystem in one platform. For developers, integrators, and edge AI teams, this means you are not just choosing hardware performance, but a complete path from prototyping to real-world deployment.

Explore Jetson Hardware

Explore NVIDIA Jetson hardware for robotics, computer vision, industrial edge AI, and embedded development. Select up to three featured products in the theme editor, or open the full NVIDIA collection for more options.

View More Products
📚 Technical Documents & Drivers
🤖
NVIDIA Jetson Platform

Explore NVIDIA Jetson for robotics, edge AI, developer kits, modules, software, downloads, ecosystem resources, and support.

Explore
🧰
NVIDIA JetPack SDK

Get the official Jetson software stack for building AI-powered edge applications with CUDA, TensorRT, Jetson Linux, and developer tools.

View SDK
⚡
NVIDIA CUDA

Access NVIDIA’s accelerated computing platform for GPU programming, CUDA Toolkit resources, libraries, documentation, and learning paths.

Learn CUDA
🚀
NVIDIA TensorRT

Optimize and deploy high-performance deep learning inference with TensorRT runtimes, compilers, model optimization, documentation, and downloads.

Explore
🦾
NVIDIA Isaac ROS

Use NVIDIA-accelerated ROS 2 packages, reference workflows, tutorials, and repositories for autonomous robots on Jetson and NVIDIA platforms.

View Docs
👁️
NVIDIA Metropolis

Build, deploy, and scale video analytics AI agents and vision AI applications from edge devices to cloud infrastructure.

Explore
🧩
NVIDIA Jetson Modules

Compare Jetson module families, performance tiers, ecosystem resources, developer kits, software, and production module options.

Compare
📅
Jetson Product Lifecycle

Check Jetson module availability timelines, lifecycle status, end-of-life notices, and long-term production planning information.

Check Lifecycle

Frequently Asked Questions

Everything you need to know about the network, hardware, and setup.

Can I start with Jetson if I am new to edge AI and robotics?

Yes. Jetson is suitable for both first-time developers and more advanced teams, especially if you start with an entry-level platform and focus on one clear use case such as computer vision or robotics prototyping.

What else do I need besides a Jetson board to start building?

In most projects, you will also need power, storage, cooling, and the right peripherals such as cameras, sensors, or networking accessories depending on your application.

Can I use Jetson for production deployment, or is it mainly for development?

Jetson can be used for both. Many teams start with development boards for prototyping and later move to embedded or deployment-ready systems for real products and field installations.

Do I need to learn the full NVIDIA software stack before getting started?

No. Most users begin with JetPack and only add tools like CUDA, TensorRT, or Isaac ROS as their project becomes more advanced.

How much power and thermal planning should I consider for a Jetson project?

It depends on the performance tier and workload. Higher-performance Jetson platforms usually require more attention to cooling, enclosure design, and power stability, especially in robotics and industrial deployments.

Do you support B2B projects, bulk purchasing, or deployment guidance for Jetson-based systems?

Yes. We support B2B and project-based inquiries for Jetson hardware selection, bulk purchasing, and deployment planning. If you would like to place a business order or purchase with VAT exemption, please refer to our B2B Order Process.⬅️click here

Register for our newsletter

Get the latest information about our products and special offers.

Website Feedback

Help us improve OpenELAB

Found a website issue or have an idea? Tell us what would make your experience better.