reComputer J401 Nano Kit with NVIDIA® Jetson Orin™ Nano 8GB Super - Seeed Studio 100088050

Description

Development kit for local artificial intelligence processing (Edge AI), equipped with an NVIDIA® Jetson Orin™ Nano 8GB module and a reComputer J401 carrier board. The reComputer J401 Nano kit supports the development of image analysis, object recognition, robotics, and generative AI applications. In Super Mode, it delivers up to 67 TOPS of AI computing performance, using the NVIDIA Ampere™ GPU architecture and AI acceleration. Its compact design and available interfaces enable the development of standalone vision systems, model testing workstations, and mobile robot prototypes.

The computing module features a six-core Arm® Cortex®-A78AE processor operating at up to 1.7 GHz and an NVIDIA Ampere GPU with 1024 CUDA® cores and 32 Tensor Cores. It provides 8 GB of LPDDR5 memory with a bandwidth of up to 102 GB/s. Super Mode increases the platform's computing capabilities by enabling module power consumption of up to 25 W, alongside the standard 7 W and 15 W power modes. The kit includes a 256 GB NVMe SSD with NVIDIA JetPack™ 6.2 pre-installed, a WiFi/Bluetooth module, and a cooling system with a fan. The manufacturer also provides system update documentation, including support for JetPack 7.2.

The carrier board provides four USB 3.2 Type-A ports with data transfer speeds of up to 10 Gb/s, Gigabit Ethernet, HDMI® 2.1, and two MIPI CSI connectors for cameras. Additional interfaces include CAN, a 40-pin expansion header, a 12-pin control and UART header, and a USB-C 2.0 port operating in device mode. M.2 Key M and M.2 Key E slots support an NVMe SSD and a wireless connectivity module, both of which are already included in the kit. The available interfaces enable cameras, sensors, motor controllers, and other peripherals used in AI systems to be connected.

The NVIDIA Jetson platform supports computer vision models, Large Language Models (LLMs), and Vision-Language Models (VLMs) using compatible development environments. NVIDIA CUDA and TensorRT™ libraries accelerate AI inference workloads. The module can be used to develop local AI assistants, image classifiers, speech recognition systems, and applications that analyze visual information.

The reComputer AI Lab platform provides example models, deployment tools, and educational resources for Edge AI computers. It includes resources for computer vision, LLM and VLM models, and containerized applications. The availability and performance of individual models depend on the JetPack version, memory requirements, and software environment configuration.

The reComputer J401 Nano supports the NVIDIA JetPack SDK, CUDA, TensorRT, DeepStream SDK, and popular PyTorch and TensorFlow libraries. Compatibility with ROS and ROS 2 allows the module to be used as a control computer in mobile robots, autonomous vehicles, and research workstations. Its interfaces enable data from cameras and sensors to be combined with object detection, navigation, and motion planning algorithms. The kit can also be used to prototype AI systems intended for subsequent deployment in industrial equipment.

The platform is suitable for intelligent surveillance, video analysis, and event recognition systems. Vision-Language Models can be used to describe scenes, answer questions about visual content, and search video footage for specific situations. Example projects include warehouse event analysis, obstacle identification, and process monitoring.

The jetson-examples GitHub repository provides ready-to-use examples for running AI models and applications on NVIDIA Jetson devices. The available scripts simplify dependency installation and deployment of selected computer vision, robotics, and generative AI solutions. Individual projects have their own RAM, storage, and software requirements.

The kit can serve as a computer for multi-stream video analytics in surveillance systems, quality inspection, and automated object counting applications. The NVIDIA Jetson Orin Nano module supports hardware-accelerated H.265 video decoding, including a single 4K stream at 60 fps or multiple lower-resolution streams. It enables the development of applications combining image analysis with object detection and transmission of results to external systems. The actual number of video streams that can be analyzed simultaneously depends on video resolution, the selected AI model, and processing configuration.

NVIDIA DeepStream and TensorRT libraries support the development of video processing applications with GPU acceleration. The kit can be used for vehicle detection and counting, object identification on production lines, and traffic analysis in public spaces. Integrating computer vision algorithms with hardware interfaces allows detection results to be transmitted to other devices and control systems.

The manufacturer provides the reComputer Jetson for Beginners course, covering Linux fundamentals, NVIDIA JetPack, CUDA and TensorRT, computer vision, and robotics. The device can also be configured and managed using the Seeed Jetson Develop Tool, which provides a graphical interface for firmware installation, diagnostics, remote management, and application deployment.

The kit uses a carrier board measuring 100 x 80.6 x 32.5 mm and weighing approximately 168 g, with mounting accessories for workstation installation. The manufacturer specifies an operating temperature range of -10°C to 60°C. Power is supplied through a DC 5.5/2.5 mm barrel connector with an input voltage of 9 V to 19 V. The kit includes a 19 V / 4.74 A power adapter, AC power cords with different plug standards, and mounting hardware. Additional information on hardware configuration, available interfaces, and system setup is provided in the reComputer J30/40 documentation on the Seeed Studio Wiki.

Features

• Computing module specifications
  • Manufacturer: Seeed Studio
  • Model: reComputer J401 Nano Bundle
  • Manufacturer part number: 100088050
  • Product type: Edge AI development kit
  • Module: NVIDIA® Jetson Orin™ Nano 8GB
  • Operating mode: Super Mode
  • AI performance: up to 67 TOPS in Super Mode
  • Processor: 6-core Arm® Cortex®-A78AE
  • Maximum CPU frequency: 1.7 GHz
  • GPU architecture: NVIDIA Ampere™
  • CUDA® cores: 1024
  • Tensor Cores: 32
  • Maximum GPU frequency: 1020 MHz
  • RAM: 8 GB LPDDR5
  • Memory bus: 128-bit
  • Memory bandwidth: up to 102 GB/s
  • Module power modes: 7 W / 15 W / 25 W
  • Pre-installed software environment: NVIDIA JetPack™ 6.2
  • System updates supported according to manufacturer documentation
  • JetPack 7.2 support according to the manufacturer
• Storage and video processing
  • SSD: 256 GB NVMe
  • SSD interface: M.2 Key M PCIe
  • Supported SSD form factor: M.2 2280
  • SSD with pre-installed operating system: included
  • Hardware H.265 video decoding: 1x 4K at 60 fps
  • Hardware H.265 video decoding: 2x 4K at 30 fps
  • Hardware H.265 video decoding: 5x 1080p at 60 fps
  • Hardware H.265 video decoding: 11x 1080p at 30 fps
  • 1080p video encoding at 30 fps using CPU cores
• Interfaces and connectivity
  • Carrier board: reComputer J401 Carrier Board
  • Ethernet: 1x RJ45 Gigabit Ethernet 10/100/1000 Mb/s
  • USB: 4x USB 3.2 Type-A, up to 10 Gb/s
  • USB: 1x USB-C 2.0 in device mode
  • Display output: 1x HDMI® 2.1
  • Camera connectors: 2x MIPI CSI, 15-pin, 2-lane
  • CAN interface: 1x
  • Expansion header: 1x 40-pin
  • Control and UART header: 1x 12-pin
  • M.2 Key M slot for SSD
  • M.2 Key E slot for WiFi/Bluetooth module
  • WiFi/Bluetooth module: installed
  • WiFi antenna kits: 2 pcs
  • Fan connector: 4-pin, 1.25 mm, 5 V PWM
  • RTC connector: 2-pin
  • RTC battery socket: CR1220
  • CR1220 RTC battery: not included
• Software and applications
  • NVIDIA JetPack™ SDK support
  • NVIDIA CUDA® and TensorRT™ support
  • NVIDIA DeepStream SDK support
  • PyTorch and TensorFlow library support
  • ROS and ROS 2 compatibility
  • NVIDIA Isaac™ tools supported in compatible configurations
  • Access to AI deployment examples through reComputer AI Lab
  • Access to projects in the jetson-examples repository
  • Access to the reComputer Jetson for Beginners course
  • Configuration using Seeed Jetson Develop Tool
  • Designed for computer vision and object recognition
  • Generative AI application development
  • Suitable for mobile and autonomous robots
  • Development of surveillance and video analytics systems
• Power supply and construction
  • Carrier board supply voltage: 9 V to 19 V DC
  • Power connector: DC 5.5/2.5 mm
  • Included power adapter: 19 V / 4.74 A
  • Board dimensions with module and cooling: 100 x 80.6 x 32.5 mm
  • Weight: approximately 168 g
  • Operating temperature range: -10°C to 60°C
  • Cooling: heatsink with fan
  • Workstation mounting using the included standoffs
  • Certifications: CE, FCC, RoHS

Kit Includes

• 1x Seeed Studio reComputer J401 carrier board
• 1x NVIDIA® Jetson Orin™ Nano 8GB module
• 1x 256 GB NVMe SSD with pre-installed operating system
• 1x heatsink with fan
• 1x WiFi/Bluetooth module
• 2x WiFi antenna kits
• 1x 19 V / 4.74 A power adapter with DC 5.5/2.5 mm connector
• 3x AC power cords with US, EU, and CN plugs
• 5x copper standoffs
• 5x mounting screws