TAIWAN — A provider of high-performance GPGPU (General Purpose Graphics Processing Unit) and AI computing solutions, Aetina Corporation, released its AN810-XNX edge AI computer that leverages the capabilities of the nVidia Jetson Xavier NX SoC (System on Chip). This AI processing board is available in several I/O configurations that are designed for things like drone, UAV, and industrial inspection, deep learning, and medical imaging.
The AN810-XNX has a Nano-ITX form factor and is composed of an NVIDIA Jetson Xavier NX board on Aetina’s new AN810 carrier board. It has a full M.2 slot which is ‘M’-keyed, another than is ‘B’-keyed. It includes support for PCI Express, SATA, USB 3.2 Gen2, and of course, USB 2.0. The AN810-XNX SBC includes support for 4G and 5G modules. The board has a built-in SIM card slot.
The AN810-XNX includes a 120-pin connector that fully supports the MIPI CSI-2 (Mobile Industry Processor Interface Camera Serial Interface) interface. The board can manage some rather intensive AI workloads like using ultra-high-resolution cameras for image analysis.
The Jetson Xavier NX’s CPU is capable of 21 TOPS (Trillion Operations per Second) which is great for running modern AI workloads. It is able to do so while consuming just 10 watts of power. The NX has the processing power to run several neural networks in parallel while processing data coming in from multiple high-resolution sensors.
“AIoT environment is diverse and complex. Throughout the edge AI systems, GPU-based platform has the most flexibility and Azure cloud services enhance the overall performance. Benefits the customer with shorter development period,” said Joe Lo, general manager of Aetina. AN810-XNX, high compatibility platform with well-backup service from edge to cloud, could deploy high-quality, software-defined features on embedded and edge devices.