25 mai 2026

Innodisk Launches High-Speed 10GbE LAN Series to Power Next-Generation Edge AI Networking

The new LAN portfolio combines compact form factors, wide-temperature support, and high-speed Ethernet performance for embedded and space-constrained AI systems.

Innodisk, a leading global AI solution provider, today launched its High-Speed 10GbE LAN Series, a portfolio of LAN modules built to address the growing networking demands of edge AI applications. As edge AI applications increasingly rely on real-time data exchange between sensors, edge servers, and endpoints, network connectivity has become a critical foundation for system performance. The series comes in both M.2 and PCIe form factors, delivering high-throughput, low-latency connectivity with DPDK, PTP, and SR-IOV, enabling efficient processing for data-intensive workloads in space-constrained environments.

Unleash Full-Speed Networking

As edge AI adoption accelerates, traditional 1GbE connectivity is increasingly unable to support real-time data processing, high-resolution video streaming, and multi-sensor integration. Innodisk’s 10GbE LAN Series tackles these challenges with low-latency connectivity powered by Intel E610/X710 Ethernet controllers, and supports DPDK 25.07/DPDK 16.11, PTP, and SR-IOV to accelerate packet processing, ensuring time synchronization, and improving resource utilization in virtualized environments. This enables consistent, high-performance networking across edge AI applications, including inference, AGV and AMR coordination, smart factory vision, and NVR surveillance.

Versatile Design, Rapid Deployment

The series is available in M.2 2242, M.2 2280, and PCIe low-profile form factors, making it easy to integrate into a wide range of embedded systems without major redesign. Its daughterboard architecture, paired with high-speed shielding cables, enables flexible installation and upgrades in space-constrained environments while reducing integration complexity and accelerating time-to-market.

Selected models also support wide-temperature operation from -40°C to 85°C, ensuring stable performance in demanding conditions, such as outdoor deployments and factory automation.

Industry Firsts in M.2 Form Factor

Innodisk expands the boundaries of M.2 networking with multiple industry-first designs. Its EGPL-T203, dual-port 10GbE LAN module with wide-temperature support, is the first of its kind in an M.2 form factor rated for -40°C to 85°C Ta, enabling reliable deployment in harsh environments. The series also introduces EGPL-T2F1, the first M.2-based SFP+ LAN module, recognized with the Embedded World 2026 Best in Show Award, supporting both optical and direct-attached copper SFP+ modules for high-speed fiber connectivity.

Alongside these innovations, additional models, including EGPL-T103, ELPL-T101, and ELPL-T201, complete the lineup, offering flexible configurations to address deployment needs. As edge AI continues to evolve, the need for faster, adaptable networking is becoming essential. Innodisk’s High-Speed 10GbE LAN Series brings together performance, compact form factors, and deployment flexibility, making it easier to build efficient, high-speed systems at the edge.

For more information, please visit https://www.innodisk.com/en/index
www.innodisk.com

Architect Intelligence – About Innodisk

Established in 2005 in Taiwan, Innodisk is a global leader in industrial-grade memory and storage solutions, maintaining the largest market share in industrial-grade storage and ranking among the top 10 worldwide providers of memory modules. Building on this strong foundation, we are now evolving beyond industrial market to pioneer innovations in Edge AI domains.

From storage to cutting-edge AI applications, our solutions serve as reliable building blocks for AI deployments, enabling real-time data sensing, accelerated computing, and intelligent decision-making at the edge. Through hardware-software integration, we create comprehensive solutions that address the dynamic needs of modern enterprises. As we expand our global presence, our goal is to accelerate the widespread adoption of AI and edge computing, contributing to architecting a more intelligent world.