New
NVIDIA Jetson AGX Orin Module

NVIDIA Jetson AGX Orin Module

NVIDIA Jetson AGX Orin Module
NVIDIA | A-000000-08130
€1,728.00
20 perc. French VAT inc.

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

Quantity
Contact us to find out about delivery times

Create a cart and turn it into a quote instantly!
Payment methods (see T&C)
Debit card
PayPal
Payment by invoice
Wire transfer

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.

    ModuleGPUCPUAI PerformanceMemoryPower Consumption
    Jetson Xavier NXVolta – 384 CUDA6-core ARM21 TOPS8 / 16 GB10 – 20 W
    Jetson Orin NXAmpere – 1024 CUDA6 / 8-core ARM70 – 100 TOPS8 / 16 GB10 – 25 W
    Jetson AGX Orin 32 GBAmpere – 1792 CUDA8-core Arm Cortex-A78AE200 TOPS32 GB15 – 40 W
    Jetson AGX Orin 64 GBAmpere – 2048 CUDA12-core Arm Cortex-A78AE275 TOPS64 GB15 – 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

    Resources for the NVIDIA Jetson AGX Orin board

    Customers who viewed this item also viewed