Alpha Networks | A Qisda Group Company Alpha Networks to Showcase 1.6T Liquid-Cooled Switch at the 2026 OCP Global Summit
Bringing Together Qisda Group Technologies to Advance AI Data Center Infrastructure
Alpha Networks (TWSE: 3380) of Qisda Group (TWSE: 2352), a leading networking equipment manufacturer, will participate in the 2026 OCP Global Summit under the theme “AI Infrastructure ORV3 OCP Rack Solutions.” Aligned with Qisda Group’s strategic focus on AI Infrastructure, Alpha Networks will showcase its 64-port, 2OU 1.6T Ethernet switch for AI data centers, alongside AI data center servers from AEWIN Technologies and high-efficiency cooling systems from GrAndvance Technology, presenting an integrated ORV3 rack solution encompassing compute, networking, and cooling. As AI computing capacity continues to expand, the company is addressing global AI data centers’ requirements for high bandwidth, low latency, and energy efficiency through high-speed networking and efficient cooling technologies, demonstrating its R&D and system integration capabilities for next-generation AI infrastructure.
As AI clusters grow in scale, data center challenges extend beyond the computing performance of individual devices to the efficiency of data exchange across large numbers of compute nodes, as well as the cooling and operational pressures associated with high-density deployments. With 1.6T high-speed networking, direct liquid cooling, and a modular architecture at the core of its showcase, Alpha Networks helps customers overcome network transmission bottlenecks while maintaining scalability and operational efficiency, providing critical networking support for next-generation AI infrastructure.
1.6T High-Speed Connectivity Supports Large-Scale AI Cluster Expansion
The 64-port, 2OU 1.6T switch on display is powered by the Broadcom Tomahawk 6 (TH6) Ethernet switch chip, delivering 102.4 Tbps of switching capacity and 1.6 Tbps connectivity per port. It supports the high-bandwidth, low-latency networking required for both scale-up (intra-rack and accelerator interconnects) and scale-out (inter-rack back-end network expansion) in AI data centers. By increasing data transfer capacity between compute nodes, the switch addresses the high-speed interconnect requirements of large AI workloads, enabling network performance to keep pace with growth in computing capacity.
For platform flexibility, the product features a modular CPU architecture that can be configured with AMD or Intel platforms according to customer requirements. It also supports the SONiC open-source network operating system and a baseboard management controller (BMC) architecture, giving customers flexibility to integrate and develop data center networking applications for different deployment environments and system management requirements.
Direct Liquid Cooling and Intelligent Monitoring Improve Cooling and Operational Efficiency
To address the cooling requirements of high-performance switch chips, Alpha Networks incorporates direct liquid cooling (DLC) technology, using liquid to remove heat from key components while reducing associated energy consumption and noise. The product uses an ORV3-standard rack format and a modular mechanical design, with quick-disconnect couplings in the liquid cooling system to simplify equipment integration and maintenance.
In addition, the switch features a BMC management module based on an ASPEED chip, integrating leak detection, equipment status monitoring, and remote management. These functions help operations and maintenance personnel track system conditions and identify anomalies early, improving management efficiency and reducing their workload. From high-speed transmission and cooling design to equipment management, Alpha Networks addresses the long-term operational requirements of AI data centers through a comprehensive product architecture.
Combining Group Technologies to Demonstrate AI Infrastructure System Integration
The showcase brings together Alpha Networks’ high-speed switch, AEWIN Technologies’ AI data center servers, and GrAndvance Technology’s cooling solutions in an ORV3 rack to demonstrate integrated compute, networking, and cooling capabilities. It highlights the strengths of technological collaboration among Qisda Group companies and enables customers to evaluate deployment approaches for high-density AI infrastructure at the rack level.
With more than 20 years of experience in networking equipment R&D and original design manufacturing (ODM), Alpha Networks continues to strengthen its integration capabilities across high-speed switching, liquid cooling, and open networking platforms, providing comprehensive support from product design through manufacturing. Moving forward, the company will continue to focus on high performance, open architectures, and energy efficiency to help customers worldwide build scalable AI data centers that support sustainability.
＃ ＃ ＃
《Alpha Networks Inc.》
Alpha Networks Inc. was founded in 2003 and is one of the largest professional network equipment manufacturers in Taiwan. With years of experience in product development and design, Alpha Networks has been a major provider of network product design, R&D, and manufacturing for renowned global brands. The Company’s product lines cover Local and Metropolitan Area Networks (LAN/MAN), wireless broadband networks, digital multimedia, and enterprise mobile solutions. Alpha Networks Inc. is headquartered in Hsinchu Science Park, Taiwan, and the Company owns 3 major R&D centers worldwide, as well as subsidiaries in North America, Japan, China, and Vietnam. For more information on Alpha Networks Inc., please refer to the company’s website (www.alphanetworks.com).
《News Contact》
Claire Chen
Tel: +886-3-563-6666 ext. 6562
E-mail: Claire_Chen@alphanetworks.com