NVIDIA Jetson Orin NX 8GB AI Edge Computing Module
The NVIDIA Jetson Orin NX 8GB is a powerful system-on-module (SOM) designed for edge AI, robotics, autonomous machines, and vision processing. Delivering up to 70 TOPS of AI performance in the smallest Jetson form factor, this compact module is ideal for applications requiring high-performance AI in space-constrained environments such as drones, handheld devices, industrial robots, and smart cameras. Built on the NVIDIA Ampere architecture with 1024 CUDA cores and 32 Tensor Cores, the Jetson Orin NX 8GB provides more than 2X the performance of Jetson AGX Xavier and 3X the performance of Jetson Xavier NX.
Key Features
-
Up to 70 TOPS AI Performance: Ampere GPU with 1024 CUDA cores and 32 Tensor Cores for accelerated AI inference
-
8GB LPDDR5 Memory: 128-bit interface with 102 GB/s bandwidth for efficient neural network processing
-
6-Core ARM Cortex-A78AE CPU: 64-bit ARMv8.2 architecture with 4MB L3 cache for heterogeneous computing
-
Compact Form Factor: Measures just 69.6 × 45 mm in a 260-pin SO-DIMM package
-
Flexible Power Modes: Configurable between 10W, 15W, 20W, and 40W (MAXN_SUPER) to balance performance and efficiency
-
7-Year Software Support: Expected support through 2030 with NVIDIA JetPack SDK and Isaac ROS
Powerful AI Computing at the Edge
The Jetson Orin NX 8GB brings AI supercomputer performance to edge devices. Its NVIDIA Ampere GPU architecture enables multiple concurrent AI inference pipelines, computer vision workloads, and deep learning acceleration. With support for CUDA 11.4+, TensorRT, OpenGL 4.6, and Vulkan 1.1, developers can deploy trained neural networks and run advanced vision algorithms efficiently. The module supports video encoding and decoding for H.265 (HEVC), H.264, VP9, and AV1 standards, making it ideal for video analytics and streaming applications.
Built for Production Robotics and Autonomous Systems
Designed as a system-on-module, the Jetson Orin NX 8GB is perfect for production deployments in robotics, industrial automation, medical imaging, smart retail, and autonomous vehicles. The module integrates seamlessly with custom carrier boards via its 260-pin connector, providing high-speed interfaces including PCIe 4.0 x4, HDMI 2.1, DisplayPort 1.4a, MIPI CSI camera inputs, USB 3.2, and Gigabit Ethernet. Its compact size and power efficiency make it suitable for battery-powered and space-constrained applications.
Complete Software Ecosystem
The Jetson Orin NX 8GB is supported by NVIDIA JetPack SDK, which includes Linux for Tegra (L4T), CUDA-X accelerated libraries, TensorRT for optimized inference, and Isaac ROS for robotics development. The module runs Ubuntu 22.04 with native ROS 2 Humble support, making it ideal for developers building autonomous robots and intelligent machines. NVIDIA provides reference AI workflows and Jetson Platform Services to accelerate time-to-market.
What's in the Box
- NVIDIA Jetson Orin NX 8GB Module (900-13767-0010-000)
- Product documentation
Full Specifications
| AI Performance |
Up to 70 TOPS (INT8) |
| GPU |
1024-core NVIDIA Ampere architecture with 32 Tensor Cores |
| CPU |
6-core ARM Cortex-A78AE v8.2 64-bit (2MB + 4MB L3 cache) |
| Memory |
8GB 128-bit LPDDR5 DRAM (102 GB/s bandwidth, 3200 MHz) |
| Video Encode |
H.265 (HEVC), H.264 |
| Video Decode |
H.265 (HEVC), H.264, VP9, AV1 |
| Display |
HDMI 2.1, DisplayPort 1.4a |
| Connectivity |
PCIe 4.0 x4, USB 3.2, Gigabit Ethernet, MIPI CSI |
| Power Modes |
10W, 15W, 20W, 40W (MAXN_SUPER) |
| Form Factor |
69.6 × 45 mm 260-pin SO-DIMM |
| Process Technology |
8nm |
| Software Support |
JetPack 6.2, Ubuntu 22.04, CUDA 11.4+, TensorRT, Isaac ROS |
Specifications are sourced from the manufacturer product page and may be updated.
