AIBOX-OrinNX Overview Specifications Wiki Technical Case Download
Buy Now DataSheet
Buy Now DataSheet
Putting Large Models into a Small Box
——AIBOX-OrinNX Flagship Large-Model AI Box
100 TOPS of computing power! Support the private deployment of mainstream large models.
Bring private AI capability to meet individual AI deployment needs.

High-performance edge computing module

NVIDIA Jetson OrinNX edge computing module (16GB version) features an octa-core ARM CPU and a 1024-core NVIDIA Ampere architecture GPU
with 32 Tensor Cores. It is capable of running multiple concurrent AI application pipelines, providing powerful inference performance.

Up to 100 TOPS of computing power

AIBOX-OrinNX delivers up to 100 TOPS of computing power and provides a 128GB PCIe NVMe SSD. The JETSON system, based on Ubuntu 22.04,
offers a comprehensive desktop Linux environment with accelerated graphics and supports libraries such as NVIDIA CUDA, TensorRT, and CuDNN.
(Support for the above content varies across different versions of the JetPack SDK. For detailed information, please refer to the NVIDIA official website.)

The private deployment of large models

Generative AI at the edge

NVIDIA Jetson Orin offers unparalleled AI compute, large unified memory, and comprehensive software stacks, delivering
superior energy efficiency to drive the latest generative AI applications. It’s capable of fast inference for any
generative AI models powered by the transformer architecture, providing superior edge performance on MLPerf.

AI software stack and ecosystem

Democratize edge AI and robotics development with the world's most comprehensive AI software stack and ecosystem, powered by generative AI at
the edge and the NVIDIA Metropolis and Isaac™ platforms. NVIDIA JetPack™, Isaac ROS, and reference AI workflows enable seamless integration
of cutting-edge technologies into your products, eliminating the need for costly internal AI resources. Experience end-to-end acceleration for AI
applications and speed your time to market using the same powerful technologies that drive data centers and cloud deployments.

Support 18-channel 1080p30 H.265 video decoding

AIBOX-OrinNX supports up to 1×8K30 (H.265), 2×4K60 (H.265), 4×4K30 (H.265), 9×1080p60 (H.265), and
18×1080p30 (H.265) video decoding, meeting the diverse demands for AI application scenarios.求

Aluminum alloy enclosure with efficient heat dissipation

The AI box features an industrial-grade all-metal enclosure with an aluminum alloy structure for thermal conduction. The side of the top cover
features a grille design for external airflow and efficient heat dissipation, ensuring computing performance and stability even under high-temperature
operating conditions.Its top cover is a porous hexagonal design, combining elegance with high efficiency. The compact, exquisite device operates
stably and meets the needs of various industrial-grade applications.

A wide range of applications

AIBOX-OrinNano is widely used in intelligent surveillance, AI education, services based on computing
power, edge computing, private deployment of large models, data security, and privacy protection.

Data security
Edge computing
Large models
AI education
Computing services
Intelligent surveillance

Specifications

AIBOX-OrinNX ( 16 GB ) AIBOX-OrinNano ( 8 GB )
Basic Specifications Module

NVIDIA Jetson OrinNX (16GB) module

NVIDIA Jetson OrinNano (8GB) module

CPU

Octa-core 64-bit Arm Cortex-A78AE v8.2 processor, up to 2.0GHz

Hexa-core 64-bit ARM Cortex-A78AE v8.2 processor, up to 1.5GHz

AI
Performance

100 TOPS

40 TOPS
GPU

1024-core NVIDIA Ampere architecture GPU with 32 Tensor Cores

Video
Encoding

H.265：1x4K60、3x4K30、6x1080p60、12x1080p30

H.265：1080p30
Video
Decoding

H.265：1x8K30、2x4K60、4x4K30、

9x1080p60、18x1080p30

H.265：1*4K60、2*4K30、5*1080p60、11*1080p30
Memory

16GB LPDDR5

8GB LPDDR5
Storage
Expansion

128GB PCIe NVMe SSD (installed inside the device)

Power

DC 12V（DC 5.5 * 2.1mm）
Power
Consumption

Normal: 7.2W (12V/600mA)

Max: 33.6W (12V/2800mA)

Normal: 7.2W (12V/600mA)

Max: 18W (12V/1500mA)
Size

93.4mm * 93.4mm * 50mm
Weight

≈ 500g
Environment

Operating temperature: -20℃～60℃,Storage temperature: -20℃～70℃, Storage Humidity: 10%～90%RH (non-condensing)

Software Support OS

The Jetson system, based on Ubuntu 22.04, offers a comprehensive desktop Linux environment with accelerated

graphics, supporting libraries such as NVIDIA CUDA, TensorRT, and CuDNN.
Large Models

Robotic models: ROS models

Large language models: The private deployment of ultra-large-scale parameter models under the Transformer

architecture, such as LLaMa2, ChatGLM, and Qwen.

Vision models: ViT, Grounding DINO, and SAM.

AI painting: Stable Diffusion V1.5 image generation model in the AIGC field.
Traditional
Network
Architectures

Traditional network architectures such as CNN, RNN, and LSTM; a variety of deep learning frameworks include TensorFlow,

PyTorch, MXNet, PaddlePaddle, ONNX, and Darknet; custom operator development.

Docker container management technology facilitates easy image deployment.
AI Software
Stack

NVIDIA Jetson Orin offers unparalleled AI compute, large unified memory, and comprehensive software stacks, delivering

superior energy efficiency to drive the latest generative AI applications. It’s capable of fast inference for any generative AI

models powered by the transformer architecture, providing superior edge performance on MLPerf.

Interfaces Ethernet

1×Gigabit Ethernet (1000Mbps / RJ45)

Video Output

1×HDMI 2.1 (4K@60fps)

USB

2×USB 3.0 (Max: 1A)

Watchdog

Support external watchdog

Other

1×Type-C (USB 2.0 OTG), 1×Console (debug serial port), 1× Recovery, 1× power button