Request a Demo

Jetson AGX Orin 64GB Module

AI Performance: Up to 275 TOPS (INT8 Sparse) for heavy edge AI workloads.
GPU: NVIDIA Ampere architecture with 2048 CUDA cores and 64 Tensor cores for deep learning acceleration.
CPU: 12-core Arm Cortex-A78AE v8.2 64-bit with 3MB L2 + 6MB L3 cache for efficient multitasking.
Memory: 64GB LPDDR5X with 204.8 GB/s bandwidth for large-scale data processing.
Power Consumption: Configurable between 15-60W, suitable for portable and industrial devices.
Connectivity: 2x NVDLA, 1x PVA v2, PCIe Gen4, and NVMe support for fast storage.
Applications: Ideal for advanced robotics, autonomous vehicles, and video surveillance systems.

Order & Purchase Details

  • In Stock
  • 100% Authentic and Guaranteed
  • Last updated: October 6, 2025

1,80 $

The NVIDIA Jetson AGX Orin 64GB Module is a flagship edge AI platform, delivering up to 275 TOPS of INT8 performance, powered by the NVIDIA Ampere GPU with 2048 CUDA cores and 64 Tensor cores. Its 12-core Arm Cortex-A78AE CPU, paired with 64GB of LPDDR5X memory at 204.8 GB/s bandwidth, ensures robust multitasking and data processing for complex AI workloads. With configurable power settings from 15W to 60W, it’s optimized for energy efficiency in demanding environments, making it a versatile choice for developers building next-generation intelligent systems.

This module excels in applications like autonomous vehicles, industrial robotics, and smart surveillance, where real-time processing of high-resolution video and large AI models is critical. Its support for 2x NVDLA engines and a Vision Accelerator (PVA v2) enhances computer vision tasks, while PCIe Gen4 and NVMe compatibility enable fast storage and data transfer. The compact form factor allows seamless integration into custom hardware, offering flexibility for OEMs and system integrators targeting high-performance edge deployments.

The Jetson AGX Orin is backed by NVIDIA’s JetPack SDK, which provides a comprehensive suite of tools like CUDA, TensorRT, and cuDNN for streamlined AI development. Developers can leverage pre-trained models from NVIDIA’s NGC catalog, ensuring rapid prototyping and deployment. Its robust software ecosystem, combined with support for ROS (Robot Operating System) and DeepStream for video analytics, makes it an ideal platform for innovators pushing the boundaries of edge AI in mission-critical applications.

Additional information

Reviews

There are no reviews yet.

Be the first to review “Jetson AGX Orin 64GB Module”

Your email address will not be published. Required fields are marked *

Related Products

Nvidia_H200_NVL_Tensor_Core_141GB_grande

NVIDIA H200 NVL PCIe 141GB HBM3e

141GB HBM3e Memory: Offers significantly increased memory capacity for large workloads. 4.8 TB/s Bandwidth: Provides exceptionally fast data transfer speeds. NVIDIA Hopper Architecture: Powers next-generation AI and HPC performance. Passive Cooling: Designed for air-cooled server environments. PCIe 5.0 x16 Interface: Ensures high-speed connectivity within servers. NVLink Support: Enables high-speed multi-GPU communication. Multi-Instance GPU (MIG): Allows for secure partitioning of the GPU. Enhanced LLM Inference: Delivers faster performance for large language models. Optimized HPC Performance: Accelerates scientific computing applications. Energy Efficient: Provides better power efficiency than older GPUs.
View More
nvidia h100 nvl 94gb gpu

NVIDIA H100 nvl 94gb gpu

Memory: Equipped with 94GB of high-speed HBM3 memory. Architecture: Based on NVIDIA's advanced Hopper architecture. Connection: Uses the high-speed PCIe Gen5 x16 interface. Cooling: Features a passive cooling system, relying on server fans. Performance: Utilizes fourth-generation Tensor Cores for acceleration. Application: Optimized for large language model (LLM) inference. Interconnect: Supports NVLink for high-speed communication between multiple GPUs. Scalability: Includes Multi-Instance GPU (MIG) capability. Power: Has a configurable power consumption of 350-400W.
View More
900-13701-0050-000-2

Jetson AGX Orin 64GB Module

AI Performance: Up to 275 TOPS (INT8 Sparse) for heavy edge AI workloads. GPU: NVIDIA Ampere architecture with 2048 CUDA cores and 64 Tensor cores for deep learning acceleration. CPU: 12-core Arm Cortex-A78AE v8.2 64-bit with 3MB L2 + 6MB L3 cache for efficient multitasking. Memory: 64GB LPDDR5X with 204.8 GB/s bandwidth for large-scale data processing. Power Consumption: Configurable between 15-60W, suitable for portable and industrial devices. Connectivity: 2x NVDLA, 1x PVA v2, PCIe Gen4, and NVMe support for fast storage. Applications: Ideal for advanced robotics, autonomous vehicles, and video surveillance systems.
View More
product-2

Jetson Orin Nano Super Developer Kit

  • High processing power: up to 70–100 TOPS
  • Compact size: ideal for space-constrained devices
  • Optimized power consumption: 10–25 watts
  • Multi-camera support: simultaneous camera connections
  • NVIDIA software compatible: Jetson Linux ready
View More
product-1

Jetson T5000 Module

AI Performance: Up to 2070 FP4 TFLOPS with Blackwell GPU for next-gen physical AI and robotics. GPU: Blackwell architecture with 1536 cores and 64 Tensor cores for large models and dense sensor data. CPU: 14-core ARM Neoverse for complex, real-time processing. Memory: 128GB LPDDR5X 256-bit with 273 GB/s bandwidth for multi-stream processing. Power Consumption: 40-130W, configurable for heavy industrial applications. Connectivity: 4x 25GbE Ethernet, 1x PVA v3, and NVMe support via PCIe Gen5. Applications: Suited for humanoid robots, autonomous systems, and high-data-volume edge AI.
View More
0
    Your Cart
    Your cart is emptyReturn to Shop

    Request Demo Form