This announcement by SynaXG, in collaboration with Kyocera and NVIDIA, represents a significant breakthrough in AI-driven wireless communications. By demonstrating the first-ever data transmission over mmWave radios using the NVIDIA GH200 Grace Hopper™ Superchip, the collaboration has taken a crucial step toward AI-native wireless networks. This achievement is more than just a technological milestone; it lays the foundation for intelligent, high-performance, and software-defined network infrastructure that will redefine the capabilities of 5G and prepare the industry for the eventual shift to 6G.
At the heart of this innovation is the integration of NVIDIA’s AI Aerial™ platform with SynaXG’s AI-RAN solution, which enables AI-based network optimization, intelligent resource allocation, and real-time data processing. By running Layer 1 to Layer 3 functions on the GH200 Superchip, this AI-RAN system offers both AI-and-RAN and AI-for-RAN capabilities, enhancing network performance while reducing operational complexity. This shift toward AI-native wireless infrastructure brings not just faster data speeds but also improvements in energy efficiency, scalability, and adaptability. The solution is fully compliant with O-RAN standards and operates as a software-defined vRAN, making it a flexible and future-proof alternative to traditional hardware-based systems.
The early laboratory trials demonstrated the impressive potential of this solution, delivering an aggregate peak throughput of up to 30 Gbps per GH200 while maintaining high energy efficiency. This performance level matches custom ASIC-based solutions but with far greater adaptability, enabling software-driven updates and upgrades without the need for costly hardware modifications. The use of AI in radio signal processing, including intelligent frequency selection and power optimization, is a game-changer in network performance, especially in enhancing cell edge throughput and reducing overall power consumption. The efficiency of mmWave technology further supports these advancements, as it enables the rapid transmission of large amounts of data with minimal energy consumption.
The impact of this technology extends far beyond telecommunications, with potential applications across multiple industries. AI-RAN’s combination of ultra-reliable connectivity and AI-driven processing is particularly suited for environments that demand real-time intelligence, such as smart manufacturing, logistics, and healthcare. In stadiums and entertainment venues, for example, AI-RAN can enable real-time video processing and live analytics, enhancing both connectivity and user experience. The collaboration between SynaXG, Kyocera, and NVIDIA is not just about advancing 5G networks; it is about shaping the future of digital ecosystems by enabling new use cases for AI at the network edge.
With trials for this AI-RAN solution scheduled for late 2025, the industry is on the cusp of a major transformation in wireless communications. The unveiling at Mobile World Congress 2025 will be a key moment for industry stakeholders to witness this innovation in action and explore its potential for real-world deployments. As operators and enterprises look for ways to optimize performance and reduce costs, AI-RAN presents a compelling vision of the future—one where AI-driven intelligence, software-defined flexibility, and high-speed connectivity converge to create smarter, more efficient networks.
Leave a Reply