Qualcomm and Arduino have teamed up to launch a new single-board computer called Arduino VENTUNO Q. This board is designed for robotics, generative AI, and edge computing, with the significant ability to work completely offline.
The VENTUNO Q features the Qualcomm Dragonwing IQ8 Series processor and a special STM32H5 microcontroller for precise control. This setup lets systems observe, make decisions, and take action all on the same device.
Qualcomm states that the board’s neural processing unit can handle up to 40 “dense TOPS” of computing power. This means it can perform complex AI tasks and processing locally and at the same time. With 16GB of RAM and 64GB of expandable storage, it can easily handle demanding multi-tasking.
Fabio Violante, VP and GM at Arduino, explained that this platform allows for “systems that don’t just interpret the world, they interact with it.” He highlighted that machines can now operate without needing to connect to the cloud.
The design of the VENTUNO Q brings AI acceleration together with real-time microcontroller logic. This combination means that sensing, deciding, and acting all happen on one single board.
VENTUNO Q can power fully independent AI agents for various tasks. These include offline voice assistants, smart mirrors that respond to gestures, and interactive kiosks found in places like transport hubs, hospitals, or tourist centers.
The system also supports advanced robotics. This includes “pick-and-place” robotic arms guided by vision and service robots that can follow people in busy areas.
Furthermore, it helps autonomous machines navigate complicated environments using Visual SLAM technology combined with optimized path planning.
Edge AI vision systems are also possible, allowing for proactive security monitoring, traffic observation, and automated quality checks using local visual language models. All these functions happen directly on the board, removing the need to send data to outside servers.
VENTUNO Q runs Ubuntu and Debian Linux on its main processor and Arduino Core on Zephyr OS for real-time control. The Arduino App Lab supports Python scripts, Arduino sketches, and ready-to-use AI models like gesture recognition and object tracking. It also supports local large language models (LLMs) powered by Qualcomm AI Hub, while Edge Impulse Studio allows users to train their own custom models.
For hardware connections, it offers industrial input/output (I/O) ports, multiple MIPI CSI camera connectors, audio, display outputs, and 2.5Gb Ethernet.
“With VENTUNO Q, AI can finally move from the cloud into the physical world,” said Fabio Violante. “This platform makes it possible to build machines that perceive, decide, and act, all on a single board. Our goal is to make advanced robotics and edge AI accessible to every developer, educator, and innovator.”
VENTUNO Q is compatible with existing Arduino UNO shields, Modulino nodes, Qwiic sensors, and Raspberry Pi HAT expansions. It will be available in the second quarter of 2026 through the Arduino Store and authorized resellers like DigiKey and Farnell. However, its exact impact on current platforms like the Raspberry Pi remains unclear.
