Jetson One Specs Explained: eVTOL Aircraft vs. NVIDIA Edge-AI Compute

The Jetson One is a $92,000 single-seat personal eVTOL aircraft, but when embedded developers and drone builders search for "Jetson One specs," they are almost always confusing the aircraft with the NVIDIA Jetson family of edge-AI system-on-modules (SoMs)—specifically the Orin and Nano series—that provide the neural processing power for autonomous flight and robotic vision. What this distinction changes in a real installation is massive: it shifts your design from wiring a simple 5V microcontroller to integrating a full Linux-based AI edge computer that requires dedicated 19V power regulation, active thermal management, and high-speed CSI camera routing. People commonly confuse the consumer "Jetson One" flying vehicle with the "Jetson Orin" or "Jetson Nano" developer kits, or mistakenly assume the aircraft's physical specs (like its 32 kWh battery or 102 km/h top speed) apply to the silicon modules.

The "Jetson One" Confusion: eVTOL Aircraft vs. Edge-AI Compute

If you are reading this on an electronics workbench, you are likely building a robot, drone, or autonomous rover. The Jetson One is a commercial personal flying vehicle manufactured by Jetson Aero in Sweden. It uses proprietary redundant flight controllers and industrial-grade compute to keep a human pilot safe in the air. You cannot buy the Jetson One's flight computer as a standalone maker component.

However, the silicon that powers the vision and path-planning systems in similar autonomous vehicles is the NVIDIA Jetson lineup. When makers talk about "Jetson specs," they are referring to the Tera Operations Per Second (TOPS), Thermal Design Power (TDP), and I/O lanes of these edge-AI modules.

NVIDIA Jetson Module Specs That Actually Matter for Makers

When sizing a compute module for a custom embedded project, ignore the marketing fluff and look at three hard constraints: AI compute (TOPS), power envelope (TDP), and physical I/O. Think of the Jetson as the cerebral cortex (handling vision and path planning) while your Pixhawk or ESP32 flight controller is the brainstem (handling raw motor PID loops and IMU stabilization).

The current generation is dominated by the Orin architecture. The Jetson Orin Nano delivers up to 40 TOPS at a tiny 7W to 15W power envelope, making it the undisputed king for battery-powered aerial and ground robots.

Worked Example: Sizing Power and Compute for an Autonomous Drone

Let's say you are building a 4kg autonomous inspection drone. You need to run a YOLOv8 neural network for obstacle avoidance alongside a visual SLAM (Simultaneous Localization and Mapping) algorithm. Your compute requirement is roughly 25 TOPS. The Jetson Orin Nano 8GB (40 TOPS) is the perfect fit, but how do you power it from your drone's flight battery?

Your drone uses a 6S LiPo battery, which has a nominal voltage of 22.2V (fully charged at 25.2V). The Jetson Orin Nano itself runs on internal 5V and 3.3V rails, but the carrier board (the breakout board the module plugs into) typically requires a 19V input via a barrel jack or XT30 connector.

1. Calculate Maximum Current: The Orin Nano maxes out at 15W TDP. Adding carrier board overhead (voltage regulators, NVMe drive, USB peripherals), your total system draw is roughly 22W.
2. Size the Buck Converter: You need a step-down (buck) converter to drop the 25.2V max LiPo voltage down to 19V for the carrier board.
   Formula: Current (I) = Power (P) / Voltage (V).
   22W / 19V = 1.15 Amps.
3. Apply Safety Margin: Always add a 20% margin for transient spikes when the AI model loads into VRAM. 1.15A * 1.2 = 1.38A.
4. Component Pick: You must select a buck converter rated for at least 2A continuous output at 19V. A standard 5V/3A BEC (Battery Eliminator Circuit) used for RC servos will not work here; you need a dedicated high-voltage switching regulator like the Pololu D24V22F19 or a custom Matek BEC-12V/19V.

Where You Meet This in Practice: Carrier Boards and Power Rails

Integrating a Jetson module changes your circuit design fundamentally. You cannot plug a Jetson into a breadboard. The raw SoM (System on Module) features high-density board-to-board connectors that mate exclusively with a carrier board.

In practice, this means dealing with three specific hardware realities:

• NVMe Storage is Mandatory: Unlike the legacy Jetson Nano which booted from a microSD card, the Orin series requires an M.2 NVMe SSD. SD cards cannot handle the IOPS (Input/Output Operations Per Second) required for Linux swap and AI model loading. You will need to source an M.2 2242 or 2280 NVMe drive (e.g., Western Digital SN570) and mount it to the underside of your carrier board.
• CSI-2 Camera Routing: Standard USB webcams introduce too much latency for real-time drone avoidance. You must use MIPI CSI-2 cameras (like the Arducam IMX477 or Leopard Imaging modules). These require delicate 15-pin or 22-pin FFC (Flexible Flat Cables). In high-vibration environments like a drone chassis, you must secure these ribbon cables with Kapton tape and locking latches, or they will rattle loose mid-flight.
• Thermal Throttling: The Orin Nano generates significant heat. If you attempt to run it without the specified aluminum heatsink and PWM-controlled fan, the silicon will hit 105°C and aggressively throttle. Your 40 TOPS will drop to under 15 TOPS within three minutes, causing your neural network frame rate to plummet from 30 FPS to 8 FPS.

Decision Tree: Which Jetson Module Should You Buy?

Use this decision matrix to select the exact hardware for your embedded project. Do not overspend on compute you cannot cool or power.

The Default Pick: For 90% of hobbyists, university robotics teams, and independent drone builders, the concrete pick is the NVIDIA Jetson Orin Nano 8GB Developer Kit (Part #945-13766-0000-000). It includes the carrier board, the heatsink, and the power supply, allowing you to flash JetPack 6.0 via USB-C and start running ROS2 nodes on day one without sourcing custom power regulators.

Frequently Asked Questions