Lanner Electronics, a global leader in network appliances and edge AI servers, today announced the upcoming launch of its most powerful Edge AI platform yet — the ECA-6050. Powered by NVIDIA Hopper GPUs, this 2U front-access server is designed specifically for enterprise and telecom infrastructure, offering unmatched compute performance, GPU scalability, and low-latency network processing for next-generation AI inference at the edge.
Features
- High GPU Density: Supports up to four 600W PCIe GPUs, including NVIDIA Hopper or NVIDIA PRO RTX 6000 Blackwell Server Edition GPUs, enabling intensive AI workloads.
- Advanced GPU Connectivity: NVIDIA NVLink facilitates high-speed GPU-to-GPU communication, ensuring parallel processing of large AI models with minimal latency.
- Multi-Tenancy Support: SR-IOV and Multi-Instance GPU (MIG) capabilities allow secure partitioning of GPU resources across multiple containers or virtual machines, ideal for AI-as-a-Service and multi-tenant inference.
- Massive Memory and Expansion: Equipped with up to 1,024GB DDR5 memory and PCIe slots for NVIDIA BlueField-3 DPUs or ConnectX-7 network adapters, enhancing network offloading and data handling.
- Optimized for AI Workloads: Supports hardware-accelerated networking and data movement offload, unlocking powerful capabilities for large language model inference and real-time video analytics.
- Enterprise-Grade Software: Compatible with NVIDIA AI Enterprise software for robust AI deployment in telecom and enterprise edge environments.
Why It Matters
The ECA-6050’s combination of GPU power, scalable architecture, and low-latency networking addresses the growing demands of edge AI applications in enterprise and telecom sectors. It is uniquely positioned to enable private generative AI, real-time video analytics, and multi-tenant AI workloads, facilitating faster and more efficient AI inference directly at the network edge.
Event Announcement
The ECA-6050 will debut globally at Computex 2025, Nangang Exhibition Center, Hall 1, Booth #I0201a. Attendees can explore Lanner’s latest innovations in edge AI technology powering the future of AI networking, telecom infrastructure, and private LLM deployments.
Quote:
“The ECA-6050 represents a strategic evolution in Lanner’s AI server portfolio, supporting up to four data center-class GPUs for high-throughput, low-latency inference,” said Jeans Tseng, CTO of Lanner Electronics. “With NVIDIA Hopper GPUs and BlueField-3 DPUs, this platform delivers the compute density and connectivity to power real-time AI and training across telecom and enterprise environments.”
