DFRobot
NVIDIA Jetson Orin Nano Super Developer Kit (67 TOPS, 8GB, 25W)
·
Bring generative AI to compact edge projects with this high-performance NVIDIA Jetson platform. It delivers up to 67 TOPS of AI performance, giving small dev...
Get notified when back in stock
Bring generative AI to compact edge projects with this high-performance NVIDIA Jetson platform. It delivers up to 67 TOPS of AI performance, giving small devices the compute needed for advanced models such as LLMs, vision transformers and vision-language models.
The kit is well suited to robotics, vision AI and other on-device AI applications where low-latency processing matters. It includes 8GB LPDDR5 memory, flexible storage options, high-speed I/O and the 40-pin GPIO header makers expect from the Jetson ecosystem.
It works with NVIDIA’s AI software stack, including TensorRT, CUDA and cuDNN, and runs Linux for Tegra (L4T) with support for NVIDIA AI SDKs. The box includes the Jetson Orin Nano Super Developer Kit ×1, a 19V power supply (45W), Type B (US, JP) power cable, Type I (CN) power cable, UPC Label, and Quick Start & Support Guide.
Features:
- Exceptional AI performance: 67 TOPS of AI performance, up from 40 TOPS, for generative models.
- Large Language Models (LLMs): Supports efficient execution of LLM workloads.
- Vision Transformers: Supports advanced vision transformer models.
- Vision-Language Models (VLMs): Supports vision-language model workloads.
- Compact & Powerful: Small form factor with industry-leading AI capabilities.
- High-Speed Memory: 8GB LPDDR5 with 102 GB/s bandwidth for seamless data handling.
- Flexible Storage: Supports SD cards and external NVMe storage for scalable data management.
- Easy Upgrade: Existing Jetson Orin Nano users can enhance performance via a simple software update.
- NVIDIA AI Ecosystem: Fully compatible with NVIDIA’s AI software stack, including TensorRT, CUDA, and cuDNN.
- AI-Driven Robotics: Suitable for AI robotics applications.
- Vision AI Applications: Suitable for vision AI applications.
Specifications:
- AI Performance: 67 TOPS (INT8)
- GPU architecture: NVIDIA Ampere architecture
- CUDA cores: 1024 CUDA cores
- Tensor Cores: 32 Tensor Cores
- CPU: 6-core Arm Cortex-A78AE v8.2 64-bit
- CPU L2 cache: 1.5MB L2
- CPU L3 cache: 4MB L3 cache
- CPU frequency: 1.7 GHz
- Memory: 8GB LPDDR5
- Memory bandwidth: 102 GB/s memory bandwidth
- Storage: SD card slot
- Storage: External NVMe support
- Power Consumption: 7W – 25W (depending on workload)
- Connectivity: Gigabit Ethernet
- Connectivity: 1x M.2 Key M (for NVMe storage)
- Connectivity: USB 3.2
- Connectivity: HDMI 2.0
- Connectivity: 40-pin GPIO header
- Operating System: Linux for Tegra (L4T) with support for NVIDIA AI SDKs
- Dimensions: 100mm × 80mm × 25mm (3.9" × 3.1" × 1.0")
- Operating Temperature: 0°C to 50°C
- Software Compatibility: Fully supports Jetson Linux and NVIDIA AI tools (TensorRT, CUDA, cuDNN, etc.)
A powerful choice for makers, students and professionals building edge AI systems, from autonomous robots to local computer vision prototypes.
Jargon buster
Plain-language definitions for the technical terms used above.
- GPIO
- General-purpose input/output pins are microcontroller pins you can set in software to read signals, switch devices on and off, or connect to peripherals. The number of GPIO pins matters because it limits how many buttons, LEDs, sensors, and other parts you can wire directly to the board.
- HDMI
- HDMI is a common digital video and audio connection used by computers, media players, and many displays. If a display kit has HDMI input, it is usually much easier to test with a single-board computer because it can act like a normal monitor.
- M.2
- M.2 is a compact edge-connector standard for plugging small modules - such as SSDs, wireless cards or microcontroller modules - into a host board without soldering. The same slot shape can carry different interfaces (for example PCIe, SATA or USB), so keying and the supported module type need to be checked.
- NVMe
- A high-speed storage standard commonly used by modern SSDs. NVMe support matters if you want faster storage for large AI models, video files or operating system images than a typical microSD card can provide.
- Tensor Cores
- Specialised processing units inside some GPUs that accelerate the matrix maths used in machine learning. They matter for choosing an AI computer because many neural-network tasks run much faster when software can use them.
- TOPS
- TOPS means trillions of operations per second, often used to describe AI accelerator performance. It helps compare whether a computing module is suited to lightweight image recognition or more demanding neural-network workloads.
Find this product in
