Taiwanese firm Aetina launched its Nvidia Jetson series of embedded systems and platforms during last week’s CES in Las Vegas.

Powered by Jetson Orin NX and Orin Nano systems-on-module (SoMs), these DeviceEdge systems and platforms are embedded computers designed for use in AI and IoT, suitable for running AI inference tasks in different verticals.

The AIB-SO21/31, AIB-SN31/41, AIB-MO22/32, AIB-MN32/42, AIE-KO21/31, AIE-KN31/41, AIE-KO22/32 and AIE-KN32/42 each have a Jetson Orin NX or Orin Nano SoM. The Orin NX modules deliver up to 100Tops of AI performance, which is up to three times the performance of the Jetson AGX Xavier module and up to five times the performance of the Jetson Xavier NX.

The Orin Nano modules provide up to 40Tops of AI performance, up to about 80 times the performance of Jetson Nano. The users of these systems and platforms can choose the SoM they want between Orin NX or Orin Nano.

The embedded system and platform models have a flexible input power range from 12 to 24V DC, can operate in temperatures from -25 to at least +70˚C, and support a data recovery and backup function. They are built with an M.2 E-Key for wifi and Bluetooth function expansion, an M.2 M-Key for storage upgrade, and an RJ-45 GbE port.

To avoid insufficient data storage, they are designed with a built-in 128Gbyte Innodisk SSD for fast data transmission.

The AIB-SO21/31, AIB-SN31/41, AIE-KO21/31 and AIE-KN31/41 use the same small-size carrier board, suitable for system integration and development with limited space; while the AIB-MO22/32, AIB-MN32/42, AIE-KO22/32 and AIE-KN32/42 are built with a median-size board to have an extra M.2 B-Key for LTE and 5G function expansion and an RJ-45 2.5GbE port for faster data transfer through an external device.

The expected available date of Orin NX-based platforms, which are the AIB-SN41, AIE-KN41, AIB-SN31, AIE-KN31, AIB-MN42, AIE-KN42, AIB-MN32 and AIE-KN32, is February 2023. The Orin Nano systems and platforms, including the AIB-SO31, AIE-KO31, AIB-SO21, AIE-KO21, AIB-MO32, AIE-KO32, AIB-MO22 and AIE-KO22, will be available in March.

The IO ports and connectors of the embedded computers are compatible with different types of sensors, allowing developers to design different AI verticals and run analytics tasks based on several kinds of data such as images, videos and audios. Moreover, Aetina offers customised carrier boards and computer cases for those who have need of any particular hardware specifications.