DEEPX, an ultra-low-power AI semiconductor company, has entered the Raspberry Pi 5 ecosystem and is targeting the physical AI developer market. By releasing an acceleration module that enables AI inference in an edge environment without server infrastructure or cloud connectivity, the company plans to provide a development path that extends from prototype development to industrial mass production.

DeepX announced on the 25th that Sixfab, an official design partner of Raspberry Pi, has globally launched the 'Sixfab AI HAT+', which was developed and manufactured by the company and equipped with its DX-M1 series NPU (Neural Network Processing Unit).

This product meets the Raspberry Pi HAT+ standard and is supplied in two performance levels: 13 TOPS and 25 TOPS.

It is scheduled to be sold through official global distribution channels with the 'Intelligentized by DEEPX' brand, DeepX's technology certification mark.

The Raspberry Pi is an open computing platform utilized for embedded systems, IoT, and AI prototype development, and is widely used by developers, makers, startups, and corporate development organizations.

Through this collaboration, DeepX They explained that they have established a point of contact to directly expand their NPU technology into this ecosystem.

It was stated that while the implementation of physical AI systems previously required dedicated AI servers and complex system integration processes, utilizing this module enables the rapid development and verification of AI applications at a relatively low cost.

This module is designed to process camera and sensor data directly within the device, reducing latency caused by cloud transmission and enabling continuous AI function execution even in unstable network environments.

It is reported that it also has strengths in terms of security because it does not transmit the original data externally.

Application fields include intelligent cameras, autonomous robots, smart agriculture, factory automation, retail analytics, and safety management systems.

Kim Nok-won, CEO of DeepX, said, “By combining the versatility of Raspberry Pi with DeepX’s ultra-low-power NPU technology, we will build an environment where developers around the world can develop and apply physical AI more easily and quickly,” adding, “We will create a physical AI ecosystem that leads from developer adoption to field verification by companies and industrial mass production.”

Roger Thornton, Director of Raspberry Pi Applications, stated, “It will provide a new path to develop and scale edge AI applications more efficiently.”