NVIDIA Jetson AGX Orin Module

Name: NVIDIA Jetson AGX Orin Module
Brand: NVIDIA
SKU: 1a8a6726-3da7-416c-9d7a-2643dbb4bb44
Price: 2718 EUR
Availability: InStock

NVIDIA | A-000000-08130

€2,718.00 VAT incl. €2,265.00 VAT excl.

The NVIDIA Jetson AGX Orin is a high-performance computing module designed for embedded artificial intelligence and Edge AI applications.

Quantity

Ask for a quote

Create a cart and turn it into a quote instantly!

Payment methods (see T&C)

NVIDIA Jetson AGX Orin – Embedded AI module (32 GB / 64 GB)

The NVIDIA Jetson AGX Orin is an embedded computing platform designed for edge artificial intelligence (Edge AI) applications. This high-end module combines an Arm Cortex-A78AE processor, an NVIDIA Ampere GPU with Tensor Cores, and dedicated AI accelerators to run deep learning models, 3D perception, and computer vision algorithms in real time.

With performance of up to 200 TOPS for the 32 GB version and 275 TOPS for the 64 GB version, the Jetson AGX Orin can process multiple video streams, handle concurrent AI applications, and run complex neural networks in embedded systems such as autonomous robots, intelligent industrial machines, advanced cameras, or multi-sensor perception systems.

Its compact 100 × 87 mm form factor makes it easy to integrate advanced AI computing capabilities into robotic platforms, industrial vision systems, or Edge AI solutions requiring high processing power.

Performance for embedded artificial intelligence

Based on the NVIDIA Orin architecture, the Jetson AGX Orin delivers a very high level of performance for embedded artificial intelligence applications. Its Ampere GPU, Tensor Cores, and deep learning accelerators enable faster inference processing, computer vision, and sensor fusion.

NVIDIA Ampere GPU with up to 2048 CUDA cores
Tensor Cores for AI inference
Arm Cortex-A78AE CPU with 8 or 12 cores depending on version
Performance up to 275 TOPS
Full support for the NVIDIA JetPack ecosystem

This architecture efficiently supports frameworks such as TensorRT, CUDA, cuDNN, Isaac, DeepStream, and NVIDIA libraries for robotics, vision, and edge generative AI.

Applications

The Jetson AGX Orin board is used in many applications requiring high AI computing power in an embedded format.

Autonomous robotics
Industrial vision
Automated inspection
Autonomous machines
AIoT and smart factories
Embedded artificial intelligence systems

Comparison of NVIDIA Jetson modules

The Jetson AGX Orin belongs to the NVIDIA Jetson family of modules designed for embedded AI. The table below compares its main features with the Jetson Orin NX and Jetson Xavier NX modules.

Module	GPU	CPU	AI Performance	Memory	Power Consumption
Jetson Xavier NX	Volta – 384 CUDA	6-core ARM	21 TOPS	8 / 16 GB	10 – 20 W
Jetson Orin NX	Ampere – 1024 CUDA	6 / 8-core ARM	70 – 100 TOPS	8 / 16 GB	10 – 25 W
Jetson AGX Orin 32 GB	Ampere – 1792 CUDA	8-core Arm Cortex-A78AE	200 TOPS	32 GB	15 – 40 W
Jetson AGX Orin 64 GB	Ampere – 2048 CUDA	12-core Arm Cortex-A78AE	275 TOPS	64 GB	15 – 60 W

A reference platform for Edge AI

The Jetson AGX Orin is a benchmark solution for integrating artificial intelligence into complex embedded systems. Combined with the NVIDIA JetPack software platform and the CUDA ecosystem, it enables the development of advanced robotics, computer vision, autonomy, and multi-sensor processing applications.

Technical specifications of the NVIDIA Jetson AGX Orin board

GPU architecture: NVIDIA Ampere
GPU:
- 32 GB: 1792 CUDA cores + 56 Tensor Cores
- 64 GB: 2048 CUDA cores + 64 Tensor Cores
AI accelerators: 2 NVIDIA Deep Learning Accelerator v2.0 (NVDLA) engines
Vision accelerator: PVA v2.0 / 7-way VLIW processor depending on version
CPU:
- 32 GB: 8-core Arm Cortex-A78AE v8.2 64-bit
- 64 GB: 12-core Arm Cortex-A78AE v8.2 64-bit
CPU cache:
- 32 GB: 2 MB L2 + 4 MB L3
- 64 GB: 3 MB L2 + 6 MB L3
AI performance:
- 32 GB: up to 200 TOPS
- 64 GB: up to 275 TOPS
Memory:
- 32 GB LPDDR5
- 64 GB LPDDR5
Memory bus: 256-bit
Memory bandwidth: up to 204.8 / 205 GB/s
Storage: integrated 64 GB eMMC 5.1
Power consumption:
- 32 GB: configurable modes 15 W / 30 W / 40 W
- 64 GB: configurable modes 15 W / 30 W / 50 W, up to 60 W
Power input: 5 V, 7 V – 20 V
PCI Express: 22 PCIe Gen4 lanes
PCIe configuration: up to 2 x8, 1 x4, 2 x1
Cameras:
- 16 MIPI CSI-2 lanes
- up to 16 virtual channels depending on configuration
- D-PHY 2.1 up to 40 Gbps / C-PHY 2.0 up to 164 Gbps
Video encoder:
- 32 GB: 1× 4K60, 3× 4K30, 6× 1080p60, 12× 1080p30
- 64 GB: 2× 4K60, 4× 4K30, 8× 1080p60, 16× 1080p30
- H.264 / H.265 / AV1 support
Video decoder:
- 32 GB: 1× 8K30, 2× 4K60, 4× 4K30, 9× 1080p60, 18× 1080p30
- 64 GB: 1× 8K30, 3× 4K60, 7× 4K30, 11× 1080p60, 22× 1080p30
- H.264 / H.265 / VP9 / AV1 support
Networking: 1 GbE RGMII + 4 × 10GbE XFI
Connector: 699-pin connector
Dimensions: 100 × 87 mm