June 2, 2026 - Santa Clara, CA and Nanjing, China

Radxa, a leader in open-source hardware and software for single-board computers and embedded systems, recently announced two new AI NAS systems based on Qualcomm's Snapdragon compute platform at an AI developer event held May 30 in Shenzhen. The new systems include DragonBay NAS, expected in September 2026, and DragonStation NAS, expected in October 2026.

The platforms highlight an important shift in the NAS market: systems are moving beyond basic storage and backup toward local AI processing. AI NAS platforms are now being designed to organize large photo and video libraries, enable intelligent search, support local inference, accelerate media workflows, and make high-performance storage more useful across home, creator, and small-business networks.

Aeonsemi’s AS22010 10GBASE-T Ethernet PHY enables dual 10GBASE-T connectivity in Radxa’s AI NAS design, providing high-bandwidth RJ45 copper Ethernet for moving large media files, accessing shared storage, and supporting AI-assisted workflows across the local network all over standard low-cost RJ45 based networking infrastructure, including widely deployed Cat 6a copper cabling up to 100 meters.

This design is a strong real-world example of the application focus behind Aeonsemi’s recently launched second-generation ChronoPHY™ AS22010 family. AS22010 was built for compact, power-sensitive platforms that need high-performance 10G Ethernet connectivity, including AI NAS systems, desktop workstations, embedded compute platforms, and edge AI infrastructure.

As more OEMs develop AI-enabled NAS and workstation-class systems, 10GBASE-T is becoming a practical requirement for local AI infrastructure.

Local AI needs local bandwidth — and AS22010 helps deliver it.

About Aeonsemi

Aeonsemi is a fabless semiconductor company building the timing and networking silicon behind modern AI infrastructure. Its portfolio includes the ChronoPHY™ multi-rate 10G Ethernet transceiver series for broadband access, enterprise networking, and Wi-Fi backhaul; the Nemo multi-gigabit automotive Ethernet (MGBASE-T1) chipset family designed for Physical AI applications including robots, autonomous vehicles, and UAVs; and the Arcadium™ family of silicon oscillators, TCXOs and synchronization timing devices. For more information, visit www.aeonsemi.com.

Contact
Mike Hansen, Senior Product Manager