en.Wedoany.com Reported - On June 3, Lite-On Technology from Taiwan, China, demonstrated its integrated AI infrastructure capabilities spanning from cloud data centers to edge computing, 5G networks, and smart city applications during COMPUTEX 2026. The exhibition covered areas including AI data center infrastructure, edge computing, 5G AI-RAN, intelligent surveillance, and traffic perception, highlighting the company's system integration capabilities across computing power, electricity, thermal management, networking, and application scenarios.
One of the key highlights of Lite-On Technology's exhibition was the open O-RU small cell solution for edge connectivity scenarios. This solution integrates the 5G AI-RAN platform based on NVIDIA AI Aerial and is equipped with an image recognition application developed by the Singapore University of Technology and Design, used to verify the collaborative operation of mobile networks, edge computing, and visual recognition tasks on the same platform. Unlike traditional communication equipment that primarily serves connectivity, AI-RAN emphasizes the integration of radio access networks with AI computing resources, enabling base stations to function not only as transmission nodes but also to support low-latency visual analysis, urban perception, industrial site recognition, and enterprise edge applications. For operators, equipment vendors, and vertical industry customers, such solutions help extend 5G networks from mere bandwidth services to computable, perceptible, and schedulable edge infrastructure.
On the data center side, Lite-On Technology simultaneously showcased products such as liquid-cooled 800V DC power cabinets, 110kW power shelves, and 280kW rack-level liquid cooling distribution units, targeting the construction needs of next-generation high-density AI servers and megawatt-level AI factories.
This product portfolio reflects the ongoing evolution of information and communication infrastructure toward the integration of "computing power, electricity, networking, and edge terminals." As AI applications scale up, data centers not only require more powerful GPUs and servers but also more efficient power architectures, more stable backup battery units, stronger thermal management capabilities, and lower-latency edge access. 5G AI-RAN has entered the industry's spotlight against this backdrop: on one hand, it brings communication networks closer to AI application sites, providing real-time data processing capabilities for factories, campuses, traffic intersections, public safety, and smart cities; on the other hand, it demands a higher level of collaboration among wireless equipment, edge servers, AI models, sensors, and cloud platforms. By placing O-RU small cells, AI-RAN platforms, intelligent surveillance, and data center power and thermal management products in the same exhibition system, Lite-On Technology demonstrates that the competitive boundaries for communication equipment suppliers are expanding. In the future, customers may no longer purchase individual hardware but rather a set of infrastructure modules capable of supporting AI deployment.
Lite-On Technology's subsidiary LEOTEK also showcased the RenAI intelligent lighting network platform and the Interlux next-generation AI traffic signal system. Interlux integrates radar and camera sensor fusion, the NVIDIA Jetson Orin edge AI inference platform, and C-V2X communication capabilities, providing low-latency processing for intersection perception and traffic decision-making scenarios. As the demand for real-time perception in smart cities and industrial sites increases, the coupling between edge AI, 5G communication, and urban infrastructure will continue to deepen. Subsequent deployment progress will depend on the pace of operator AI-RAN trials, the rollout of urban traffic projects, the cost of edge devices, and whether industry customers are willing to place visual recognition and on-site data processing capabilities at the network edge.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
