Aetina Corporation, a leading edge AI solutions provider, offers high-performance hardware and easy-to-use software platform to help AI system integrators and developers create different types of AI applications. At Computex 2022, Aetina is introducing its “Pro-AI Service” with live demonstrations of various solutions and edge AI platforms.
The Pro-AI Service, consisting of Aetina’s AI hardware and software, makes AI system development processes easier and smoother for AI developers, helping them build reliable AI application systems for use in factories, cities, retail stores, and hospitals. Aetina’s hardware includes DeviceEdge and MegaEdge that run AI inference tasks, as well as the SuperEdge AI training platform and AI accelerators. The company also provides an edge device management tool, EdgeEye that helps its users remotely monitor the real-time status of multiple edge computing devices.
Besides the Pro-AI Service, Aetina will showcase its latest AI platform products that include MegaEdge AIP-SQ37, MegaEdge AIP-FQ47, and DeviceEdge that is powered by the new NVIDIA Jetson AGX Orin. The two MegaEdge AI inference platforms support NVIDIA GPUs with MXM and PCIe interfaces respectively; both platforms have rich I/O interface designs and can be easily integrated into edge AI systems.
Aetina’s DeviceEdge with an NVIDIA Jetson AGX Orin module delivers up to 275 TOPS of AI performance for powering autonomous systems, which is eight times higher than its predecessor, NVIDIA Jetson AGX Xavier. The DeviceEdge, which can be used in out-of-band systems, has 1 x 1/10GbE PoE port, 2/4 x IEEE 802.3af GbE PoE ports, 1 x M.2 B-Key to support cellular function, 1 x M.2 E-Key to support WIFI/BT function, and 1 x M.2 M-Key for storage expansion. It also has a wide power input range of 9 to 36 VDC.
As an Elite member of the NVIDIA Partner Network, Aetina provides a wide range of edge AI platforms powered by NVIDIA Jetson modules to deliver competent performance at the edge. Additionally, customers can leverage a powerful end-to-end solution for AI application development powered by the NVIDIA TAO Toolkit, DeepStream and Triton Inference Server to easily fine-tune pretrained AI models, run PoC tasks online, and deploy improved AI models on different edge devices with NVIDIA Fleet Command. At the same time, most of the AI model improvement and deployment tasks can run on Aetina’s AI platforms. The AI platforms and NVIDIA’s development tools are a part of Aetina's demonstrations at Computex 2022.