Jetson Orin Nano
Field-Prototyping Kit

Built to kickstart AI projects with vision, audio, and LLM capabilities.

Base Platform: Jetson Orin Nano Super Developer Kit

The foundation of the Field-Prototyping Kit is the NVIDIA Jetson Orin Nano Super Developer Kit, a powerful edge AI platform delivering 67 INT8 TOPS at just $249.

Compute Performance

  • •67 INT8 TOPS AI Performance
  • •1.7× faster than predecessor
  • •Ampere GPU Architecture
  • •Tensor Cores for AI acceleration

CPU

  • •6-core Arm Cortex-A78AE
  • •Enhanced performance mode
  • •Optimized for edge AI
  • •Low-power operation

Memory

  • •8 GB LPDDR5
  • •102 GB/s bandwidth
  • •68 GB/s effective bandwidth
  • •Shared CPU/GPU memory

Form Factor

  • •100 × 79 × 21 mm
  • •Compact SoM + carrier board
  • •Integrated thermal solution
  • •Production-ready design

Connectivity

  • •4× USB 3.2 Gen 2
  • •DisplayPort 1.2
  • •M.2 Key E (Wi-Fi included)
  • •Gigabit Ethernet

Camera Interface

  • •2× MIPI CSI-2 connectors
  • •Up to 4 lanes per camera
  • •Multi-camera support
  • •Hardware ISP acceleration

Field-Prototyping Kit: Core Compute

The Field Kit builds upon the Super Developer Kit with production-ready enhancements and pre-configured software.

GPU Architecture

512-core NVIDIA Ampere GPU with 16 Tensor Cores (Field Kit configuration)

Note: Base Super Developer Kit features full 1024-core GPU; Field Kit uses optimized power profile

Storage System

High-speed NVMe SSD with pre-installed OS and complete AI stack. Significantly faster boot times, package installs, and model loading compared to SD card operation.

Memory Configuration

8 GB LPDDR5 RAM with optimized 16 GB NVMe-backed swap file. ZRAM disabled for improved stability under sustained GPU workloads.

Connectivity & Interfaces

Comprehensive connectivity options for development and deployment.

Network

  • •Gigabit Ethernet
  • •Wi-Fi/Bluetooth (configuration varies)
  • •Avahi for .local networking
  • •DNSMasq + NetworkManager tuned
  • •Hostname pre-set for LAN discovery

USB & Expansion

  • •USB 3.2 high-speed data
  • •USB-C (device mode supported)
  • •40-pin GPIO header
  • •M.2 expansion options

Imaging & Vision System

Professional-grade camera system pre-configured for computer vision applications.

Dual IMX219 Stereo Camera Module

  • •Resolution: 8 MP per camera
  • •Interface: CSI-2
  • •Pre-configured via Jetson-IO
  • •Hardware ISP support

Supported Pipelines

  • •GStreamer hardware acceleration
  • •V4L2 direct capture
  • •OpenCV CUDA-accelerated
  • •jetson-containers vision nodes

Pre-Verified Applications

Sample apps for real-time inference, object detection, depth estimation, and stereo vision demos included and tested.

Audio I/O System

Production-ready audio subsystem configured for speech and voice applications.

Microphone Array

  • •Dual microphones
  • •Pre-configured as default input
  • •ALSA + PulseAudio configured
  • •Ready for STT/VAD pipelines

Speaker Output

  • •Dual speakers
  • •Pre-configured as default output
  • •Reliable audio routing
  • •Ready for TTS immediately after boot

Operating System & Platform

Optimized Linux environment with all packages pre-configured for immediate AI development.

System Configuration

  • •OS Location: Full system runs from NVMe SSD
  • •Kernel: Updated Linux kernel with latest packages
  • •Swap: 16 GB NVMe-backed swap file
  • •ZRAM: Disabled for improved stability
  • •Hostname: Pre-set for LAN discovery
  • •Networking: Avahi for .local access

Performance Enhancements

  • •NVMe-tuned I/O for rapid model loading
  • •CUDA-F16 optimization for reduced memory footprint
  • •Verified under sustained GPU workloads
  • •System image stripped of unnecessary caches

CUDA-Accelerated Software Stack

All components pre-installed, configured, and tested. Ready for production deployment.

AI/ML Frameworks

  • •CUDA Toolchain (12.x)
  • •cuDNN
  • •TensorRT-ready containers
  • •PyTorch (JetPack-compatible)
  • •ONNX Runtime GPU

LLM / NLP Runtimes

  • •llama.cpp (CUDA-F16, compute 8.7)
  • •Ollama with GPU acceleration
  • •CTranslate2 (CUDA + cuDNN)
  • •Local-only tool-calling support

Speech Processing

  • •Whisper.cpp (CUDA)
  • •Real-time STT pipelines verified
  • •Preloaded base English models
  • •TTS via CTranslate2

Computer Vision

  • •OpenCV 4.12 (CUDA-accelerated)
  • •cuDNN DNN support
  • •GStreamer hardware pipelines
  • •Ultralytics YOLO (Jetson-optimized)
  • •Roboflow Inference Server

Containerization & DevOps

  • •Docker + NVIDIA container runtime
  • •jetson-containers ready
  • •NGINX for local APIs
  • •jetson-stats monitoring

Developer Tooling

  • •Python 3 + pip, venv
  • •GCC, g++, cmake, ninja-build
  • •Full CUDA toolchain in PATH
  • •git, wget, curl, jq

Application Layer

Pre-loaded with tested, field-ready applications and templates.

Offline Voice Assistant

Complete pipeline for local voice interaction without cloud dependencies

Multimodal Agent Templates

Pre-configured templates combining vision, audio, and LLM capabilities

Kiwix Offline Knowledge

Offline Wikipedia and knowledge base engine for air-gapped deployment

Example Camera Viewers

Ready-to-use camera preview and capture applications

Inference Demos

Sample object detection and classification demos

Development Repository

jetson-field-kit repo with examples and utilities

Media & Utilities

Essential media tools and utilities pre-installed for testing and development.

Media Tools

•VLC media player
•mpv video player
•feh image viewer
•qv4l2 camera utility
•GStreamer CLI tools
•FFmpeg (CUDA-accelerated)

Dimensions & Enclosure

Compact form factor suitable for field deployment and prototyping.

Physical Specifications

  • •Footprint similar to Jetson Dev Kit + mounted peripherals
  • •Ports accessible for field use
  • •Optional case files (STEP) available for modification
  • •Varies based on kit enclosure selection

What's Included

Everything needed to start developing AI applications immediately.

Hardware Components

  • •Jetson Orin Nano Super Dev Kit
  • •NVMe SSD with preloaded OS + software
  • •Dual IMX219 camera module
  • •Microphones + speakers (installed & configured)
  • •Power supply

Software & Documentation

  • •Pre-configured system image
  • •Complete AI stack pre-installed
  • •Example projects and demos
  • •Documentation and setup guides
  • •Development repository access

All specifications subject to change. Performance may vary based on configuration, workload, and environmental conditions. NVIDIA, Jetson, CUDA, and TensorRT are trademarks of NVIDIA Corporation.