A breakthrough in embedded applications.
The Jetson Xavier™ family of modules features the world’s first computers designed specifically for autonomous machines. Up to 32 TOPS of AI performance makes these modules ideal for handling visual odometry, sensor fusion, localization and mapping, obstacle detection, and path-planning algorithms that are critical to next-generation robots. With five different production modules based on the same architecture, theJetson Xavier family delivers just what you need to start developing for your specific robotics and edge AI applications.
Jetson Xavier sets a new bar for compute density, energy efficiency, and AI inferencing capabilities on edge devices. It’s the next evolution in intelligent machines with end-to-end autonomous capabilities.
Jetson AGX Xavier-series modules feature a compact form factor of just 100mm x 87mm and 750Gdps of high-speed I/O, delivering workstation performance at 1/10 the size. Ideal for autonomous machines, it has configurable power profiles and is available in 64GB, 32GB, and Industrial versions.
Jetson Xavier NX-series modules deliver up to 21 TOPS of AI performance in a small form factor measuring only 70mm x 45mm. They run multiple neural networks in parallel and support high-resolution sensors, making them perfect for embedded systems. With high performance, power efficiency, and support for all popular AI frameworks, these modules are production ready and available in 16GB and 8GB versions.
Jetson AGX Xavier series | Jetson Xavier NX series | ||||
Jetson AGX Xavier Industrial | Jetson AGX Xavier (64GB) | Jetson AGX Xavier (32GB) | Jetson Xavier NX (16GB) | Jetson Xavier NX (8GB) | |
AI Performance | 30 TOPS | 32 TOPS | 21 TOPS | ||
GPU | 512-core NVIDIA Volta architecture GPU with 64 Tensor Cores | 384-core NVIDIA Volta™ architecture GPU with 48 Tensor Cores | |||
GPU Max Frequency | 1211 MHz | 1377 MHz | 1100 MHz | ||
CPU | 8-core NVIDIA Carmel Arm®v8.2 64-bit CPU 8MB L2 + 4MB L3 |
6-core NVIDIA Carmel Arm®v8.2 64-bit CPU 6MB L2 + 4MB L3 |
|||
CPU Max Frequency | 2.0 GHz | 2.2 GHz | 1.9 GHz | ||
DL Accelerator | 2x NVDLA | 2x NVDLA | |||
DLA Max Frequency | 1.2 GHz | 1.4 GHz | 1.1 GHz | ||
Vision Accelerator | 2x PVA | 2x PVA | |||
Safety Cluster Engine | 2x Arm® Cortex®-R5 in lockstep | - | - | ||
Memory | 32GB 256-bit LPDDR4x (ECC support) 136.5GB/s |
64GB 256-bit LPDDR4x 136.5GB/s |
32GB 256-bit LPDDR4x 136.5GB/s |
16GB 128-bit LPDDR4x 59.7GB/s |
8GB 128-bit LPDDR4x 59.7GB/s |
Storage | 64GB eMMC 5.1 | 32GB eMMC 5.1 | 16GB eMMC 5.1 | ||
Video Encode | 2x 4K60 (H.265) 6x 4K30 (H.265) 12x 1080p60 (H.265) 24x 1080p30 (H.265) |
4x 4K60 (H.265) 8x 4K30 (H.265) 16x 1080p60 (H.265) 32x 1080p30 (H.265) |
2x 4K60 (H.265) 4x 4K30 (H.265) 10x 1080p60 (H.265) 22x 1080p30 (H.265) |
||
Video Decode | 2x 8K30 (H.265) 4x 4K60 (H.265) 8x 4K30 (H.265) 18x 1080p60 (H.265) 36x 1080p30 (H.265) |
2x 8K30 (H.265) 6x 4K60 (H.265) 12x 4K30 (H.265) 26x 1080p60 (H.265) 52x 1080p30 (H.265) |
2x 8K30 (H.265) 6x 4K60 (H.265) 12x 4K30 (H.265) 22x 1080p60 (H.265) 44x 1080p30 (H.265) |
||
CSI Camera | Up to 6 cameras (36 via virtual channels) 16 lanes MIPI CSI-2 D-PHY 1.2 (up to 40 Gbps) C-PHY 1.1 (up to 62 Gbps) |
Up to 6 cameras (36 via virtual channels) 16 lanes MIPI CSI-2 | 8 lanes SLVS-EC D-PHY 1.2 (up to 40 Gbps) C-PHY 1.1 (up to 62 Gbps) |
Up to 6 cameras (24 via virtual channels) 14 lanes MIPI CSI-2 D-PHY 1.2 (up to 30 Gbps) |
||
PCIe* | 1 x8 + 1 x4 + 1 x2 + 2 x1 (PCIe Gen4, Root Port & Endpoint) |
1 x4 (PCIe Gen4) + 1 x1 (PCIe Gen3) | |||
USB* | 3x USB 3.2 Gen2 (10 Gbps) 4x USB 2.0 |
1x USB 3.2 Gen2 (10 Gbps) 3x USB 2.0 |
|||
Networking* | 1x GbE | 1x GbE | |||
Display | 3 multi-mode DP 1.4/eDP 1.4/HDMI 2.0 | 2 multi-mode DP 1.4/eDP 1.4/HDMI 2.0 | |||
Other I/O | 5x UART, 3x SPI, 4x I2S, 8x I2C, 2x CAN, PWM, DMIC, GPIOs | 3x UART, 2x SPI, 2x I2S, 4x I2C, 1x CAN, PWM, DMIC & DSPK, GPIOs | |||
Power | 20W - 40W | 10W - 30W | 10W - 20W | ||
Mechanical | 100mm x 87mm 699-pin connector Integrated Thermal Transfer Plate |
69.6mm x 45mm 260-pin SO-DIMM connector |
Robotics and automation are increasingly being used in manufacturing, agriculture, construction, energy, government, and other industries. These applications often require extended temperature, shock, and vibration specifications to operate in harsh environments. The NVIDIA Jetson platform offers multiple options for rugged applications.
Learn more about Jetson Xavier and how ecosystem partners and customers are tapping into the Jetson platform to bring powerful AI solutions to market.
The Jetson partner ecosystem offers a wide range of AI and system software, developer tools, and custom software development. A broad range of Jetson Xavier systems and carrier boards are also available, as well as peripherals such as sensors, cameras, connectivity modules (5G, 4G, Wi-Fi), and more.
All NVIDIA Jetson modules and developer kits are supported by the same software stack, so you can develop once and deploy everywhere. Jetson software is designed to provide end-to-end acceleration for AI applications and accelerate your time to market. This brings the same powerful NVIDIA technologies that power data centers and cloud deployments to the edge.
Many AI applications have common needs: classification, object detection, language translation, text-to-speech, recommender engines, sentiment analysis, and more. Pretrained models from the NGC™ catalog are performance-optimized and ready to be fine-tuned with the NVIDIA TAO Toolkit and customer datasets, reducing time and cost for production of AI software development.
NVIDIA software platforms—including Isaac™ for robotics, Riva for AI speech, and Metropolis for video analytics—are available to speed vertical-specific applications.