Physical AI News (Japan)

Physical AI, robotics and compute — reported from Japan.

日本語
· Product launches· 日本語で読む

Design News spotlights Arduino’s “VENTUNO Q,” an edge AI development board

Design News featured Arduino’s “VENTUNO Q” on November 14, 2026. Arduino’s official store says it has an NPU capable of up to 40 dense TOPS and a microcontroller for real-time control. Its key feature is combining AI inference and device control on a single board.

What happened

On November 14, 2026, Design News featured Arduino’s “VENTUNO Q” edge AI development board. The product page on Arduino’s official store says it combines a Qualcomm Dragonwing IQ8 processor and an STM32H5 microcontroller, with NPU processing performance of up to 40 dense TOPS.

Why it matters

According to Arduino’s official store, the VENTUNO Q combines an AI inference processor and a microcontroller for real-time control on a single board, with a setup for controlling external motors and CAN-FD devices. For developers seeking to combine edge-side inference and motion control, the product’s distinguishing feature is that the necessary components are available on one board.

Analysis

The VENTUNO Q’s key point is not only its processing capacity of up to 40 dense TOPS, but also its dual-system design, with an inference processor and a device-control microcontroller on one board. According to the official store, the Linux side handles AI inference, while the STM32H5 side controls motors and I/O using Zephyr RTOS. The design can be understood as assigning AI decision-making and device operation to separate computing systems connected by an RPC bridge. However, the information provided does not include measured response performance, the range of supported models or power consumption, so the 40 dense TOPS figure alone is not enough to assess real-world processing performance. In a related article, we reported that Arduino had announced the Opta industrial micro PLC, developed jointly with Finder Arduino announces Opta, an industrial micro PLC developed jointly with Finder. The Opta article’s summary describes a product in the industrial PLC category, while the VENTUNO Q’s official description presents an integration of edge AI inference and real-time control. There is no basis for treating the two as products for the same use or as a connected product family, but it is clear that Arduino handles both industrial control equipment and AI-capable boards. The information available does not indicate whether the VENTUNO Q works with existing products such as Opta. On the computing side, the VENTUNO Q uses Qualcomm Dragonwing IQ8. Separately, an article by we on Qualcomm’s agreement to acquire PickNik Robotics reported plans to integrate Dragonwing and MoveIt Qualcomm agrees to acquire PickNik Robotics; MoveIt to remain open source. The two reports show Qualcomm’s computing platforms appearing in both edge devices and robotics software, but they do not mean that the VENTUNO Q works with MoveIt or that it has been deployed in a particular robot. In practice, the product’s appeal is likely to depend on supported AI models, controllable devices, the ease of using its development environment, price and shipping terms. The information provided does not include pricing, general availability, power requirements, performance measurement conditions or details of supported models. Until these are disclosed, it is not possible to assess the VENTUNO Q’s effective performance against existing edge devices or its suitability for moving from prototyping to sustained operation.

Sources

Translated from the Japanese original: /news/design-news-ventuno-q-wo-shoukai-05a44d. The Japanese version is authoritative if they differ.