NVIDIA-certified MGX modular server, tailored for Edge AI applications. ▪ Powered by Intel® Xeon® 6 with P-cores. ▪ Short and Compact Accelerated server - 420mm (16.5") short-depth. ▪ Optimal Accelerated performance at Edge - supporting 2 double deck GPUs and NVLink bridge for boosting GPU-to-GPU interconnections. ▪ Supporting NVIDIA® Bluefield-3 and NVIDIA® ConnectX®-7 for high performing networking.
Aetina is thrilled to unveil AEX-2UA1, the enterprise-level edge server. AEX-2UA1 complies with NVIDIA MGX™ architecture, delivering maximum performance within a compact form factor for edge computing. Powered by advanced technologies from Intel and NVIDIA, AEX-2UA1 supports a single socket Intel® Xeon® 6 and two double-deck GPUs with NVLink bridges, enhancing LLM (large language model) training and inferencing performance with direct GPU-to-GPU interconnect to unleash uncompromising edge computing performance. AEX-2UA1 empowers enterprises to effectively handle sensitive data with on-premise LLM and other enterprise AI applications. It is particularly valuable in sectors such as Finance and Medical, where privacy is paramount, and 5G Telco, offering compact, and optimal mission critical system. It serves as an ideal entry system for enterprises in need of LLM deployment, offering low barriers to entry compared to standard GPU servers in the market.