Arduino UNO Q: Linux Computing Meets Real-Time Control in the Classic Form Factor

If you've ever needed to combine a Linux computer with an Arduino for AI and machine learning projects, the UNO Q offers both within a single board. This latest release from Arduino pairs a Qualcomm Dragonwing QRB2210 processor running Debian Linux with an STMicroelectronics STM32U585 microcontroller, all within the standard UNO form factor.

Figure 1: The Arduino UNO Q hybrid single-board computer and microcontroller development board (top view). (Image source: Arduino)

Launched in October 2025 following Qualcomm's acquisition of Arduino, the board solves a persistent problem in Edge AI development: you need Linux for running AI models and processing power, but you also need deterministic real-time control for sensors and actuators.

Rather than connecting separate boards, the UNO Q integrates both processors on one PCB, providing a complete platform for projects spanning machine learning, computer vision, and real-time hardware interfacing.

Qualcomm Partnership and Processor Capabilities

Arduino's acquisition by Qualcomm brings industrial-grade silicon to the maker space while preserving Arduino's independence as a subsidiary. The Qualcomm Dragonwing QRB2210 sits at the heart of this design, providing a quad-core Arm® Cortex®-A53 processor operating at 2.0 GHz alongside an Adreno 702 GPU. You get 3D graphics acceleration via OpenGL ES and OpenCL, capabilities that extend well beyond typical Arduino applications.

Qualcomm originally designed this processor family for robotics and industrial IoT deployments, which means prototypes you develop on the UNO Q can scale directly to production hardware using identical silicon.

The imaging capabilities deserve particular attention for computer vision work. The QRB2210 contains dual Image Signal Processors that handle camera inputs up to 13 MP + 13 MP simultaneously, or a single 25 MP feed, all at 30 frames per second. You can process this visual data without offloading to external hardware. A dedicated low-power DSP takes care of audio tasks including keyword detection, while onboard AI acceleration handles inference workloads locally. Object recognition, anomaly detection, and similar applications run entirely on the board without requiring Internet connectivity or Cloud services.

Working alongside the Linux processor, the STM32U585 microcontroller runs at 160 MHz with 2 MB of flash memory and 786 KB of SRAM. This chip executes Arduino Core on top of Zephyr RTOS, giving you the familiar Arduino programming environment for real-time tasks. The architecture lets you run Python scripts and AI models on the Linux side while simultaneously handling time-sensitive operations like sensor sampling or motor control on the microcontroller.

Memory, Connectivity, and Physical Specifications

You can choose between two memory configurations when selecting your UNO Q. The entry-level variant ships with 2 GB LPDDR4 RAM and 16 GB eMMC storage. Arduino also offers a higher-tier model with 4 GB RAM and 32 GB storage. If you're planning to run GUI applications or work with larger AI models, you should opt for the 4 GB version.

For wireless connectivity, the board uses a WCBN3536A module supporting Wi-Fi 5 across both 2.4 GHz and 5 GHz bands, plus Bluetooth 5.1. Antennas are built directly into the board. The single USB-C connector handles multiple functions: host and device modes, power delivery, and even video output. If you want to use the UNO Q as a standalone computer, you'll need to add a USB-C hub for connecting peripherals like keyboard, mouse, and additional displays. Power comes through either the USB-C port at 5 V / 3 A, or via the barrel jack accepting a 7 V to 24 V input.

At 68.85 x 53.34 mm, the physical dimensions match the classic UNO exactly, so your existing shields will fit without modification. Beyond the standard Arduino headers, you'll find specialized high-speed connections for MIPI-CSI cameras, MIPI-DSI displays, and audio interfaces. The board includes built-in interaction elements: an 8 x 13 blue LED matrix, four RGB LEDs, and a user button. Communication protocols span the full range you'd expect: I2C/I3C, SPI, PWM, CAN, UART, and ADC. The Qwiic connector provides tool-free connections to compatible Modulino modules.

Choosing Your Development Tools and Programming Approach

When you power on the UNO Q, you'll find the Arduino App Lab already installed and ready to use. This new integrated development environment brings together Arduino sketches, Python code, and containerized AI models in one workspace. Pre-built components called ‘Apps and Bricks’ speed up common prototyping tasks. The interface supports development for both the Linux processor (MPU) and the real-time microcontroller (MCU) without switching tools.

If you prefer established workflows, you can continue using Arduino IDE 2.0 or newer versions to program the STM32 microcontroller. For Linux development, standard SSH access gives you terminal control. The Debian operating system includes Docker and Docker Compose, so you can deploy AI models in containers just as you would on any Linux system. App Lab runs on your development computer under Windows 10 and later, macOS 11 and later, Ubuntu 22.04 and later, or Debian Trixie.

Programming language options match each processor's role. Write Arduino sketches in C/C++ for the STM32 microcontroller to handle hardware control tasks with deterministic timing. Use Python on the Linux processor for application logic and computationally intensive workloads.

For AI inference, the STM32 microcontroller runs TensorFlow Lite for Microcontrollers through Arduino's library system, handling lightweight models efficiently. The Linux processor manages heavier AI workloads and comes with six pre-configured models ready to deploy: object detection, human detection, anomaly detection, image classification, sound recognition, and keyword spotting. All inference runs locally without requiring internet connectivity.

Potential Applications

Computer vision systems benefit from the dual Image Signal Processors and AI acceleration for tasks like object recognition or quality inspection, while the microcontroller precisely manages motors, servos, and other actuators. Robotics projects that combine visual navigation with motor control fit this profile perfectly. Industrial inspection systems can analyze images for defects while simultaneously controlling product handling mechanisms. Interactive installations and kiosks can process visual inputs locally without latency, and voice-controlled devices leverage the onboard keyword spotting without cloud dependencies.

The board's integration with Edge Impulse, which Qualcomm acquired in March 2025, provides your primary workflow for machine learning development. You can train custom models and deploy them to the UNO Q through Edge Impulse's platform. PyTorch isn't officially supported based on Arduino's documentation, but the underlying Debian system could theoretically run it with appropriate installation, though performance would depend on your model complexity.

The UNO Q vs. Alternative Platforms

Understanding where the UNO Q fits in the overall development board market can help you make informed purchasing decisions. The Arduino UNO R4 serves as a pure microcontroller platform focused entirely on hardware control and simple embedded tasks. The UNO Q builds on this foundation by adding full Linux computing for AI, image processing, and network applications, while keeping the real-time I/O that makes Arduino useful. You gain substantially more capability but also more complexity.

Looking at Raspberry Pi 5 boards at similar price points, you'll notice different tradeoffs. The Pi 5 delivers faster processing for compute-intensive tasks but has no integrated microcontroller for deterministic timing. If your project needs both capabilities, you'd need to buy a Raspberry Pi and an Arduino separately, then handle the interconnection yourself. The UNO Q eliminates this two-board requirement.

Select the UNO Q when your project requires integrated Linux and real-time control, when power efficiency matters for battery operation, or when Arduino shield compatibility provides value through existing hardware investments. Choose alternatives for different scenarios: ESP32 or UNO R4 when you only need microcontroller functionality, Raspberry Pi 5 when maximum computing performance outweighs real-time requirements, and ESP32 when budget constraints favor simplicity over AI capability.

Software Maturity and Current Limitations

Early adopters in Arduino's community forums report both successful projects and growing pains. Completed projects include smart speakers, weather stations, and face-tracking robots. Users specifically praise App Lab for simplifying tasks like creating web servers that control LEDs. However, you should expect some rough edges typical of new platforms. Reported issues include USB-C cables that don't establish reliable connections, compatibility problems with certain sensor libraries like DHT, and an initial bug in the Windows Serial Monitor that Arduino has since patched.

Plan for several practical requirements in your workflow. Since Wi-Fi connectivity is essentially mandatory during development and programming, the single USB-C port means you'll need a hub for standalone operation when using the board as a complete computer. You cannot run the STM32 microcontroller independently of the Linux system, so both processors operate together. Arduino continues releasing active updates and maintains responsive support in the forums. Complete hardware documentation including schematics and Gerber files is available for detailed technical reference.

Figure 2: The Arduino UNO Q hybrid single-board computer and microcontroller development board (bottom view). (Image source: Arduino)

Making the Decision for Your Next Project

The Arduino UNO Q successfully delivers enterprise-level processing from Qualcomm in a package accessible to makers and engineers. The dual-processor design directly solves the common requirement to combine single-board computers with microcontrollers for advanced projects involving AI.

If you're developing robotics systems with vision processing, industrial equipment with machine learning inspection, or IoT devices that need local intelligence paired with real-time sensor control, the UNO Q provides these capabilities on one board.

For more information please visit Arduino UNO Q.