Core Insights - SynaXG has achieved significant advancements in AI-native radio access networks, demonstrating the capability for concurrent operation of 5G FR1, 5G FR2, and AI workloads on NVIDIA's AI-RAN platform, confirming readiness for commercial deployment [1][9] Group 1: Technical Achievements - SynaXG attained carrier-grade 5G FR1 performance on a single NVIDIA GH200 platform, operating 20 x 100MHz 5G NR cells with over 36Gbps aggregated throughput and sub-10 millisecond latency, supporting up to 1,200 connected user devices per cell [2] - The company achieved the world's first carrier-grade FR2 virtualized RAN implementation running concurrently with FR1 and AI workloads, delivering high throughput and end-to-end latency as low as 5 milliseconds, suitable for dense urban and enterprise environments [3] - The AI-RAN system operated continuously under sustained load, validating its stability, scalability, and operational resilience for carrier environments [5] Group 2: Software and Infrastructure - SynaXG's real-time orchestration software dynamically allocates GPU resources based on live RAN and AI key performance indicators, ensuring deterministic RAN performance while maximizing infrastructure utilization [4] - The AI-RAN platform is built on a software-defined architecture leveraging NVIDIA AI Aerial platforms, supporting deterministic network performance alongside AI processing on shared GPU infrastructure [6] Group 3: Future Developments - SynaXG plans to extend its AI-RAN platform to additional NVIDIA platforms, including NVIDIA DGX Spark, showcasing the portability of its software-defined architecture [7] - The company is advancing AI-for-RAN capabilities focused on network optimization, automation, and energy efficiency, collaborating with partners to enhance performance and operational simplicity in AI-native networks [8] Group 4: Industry Impact - SynaXG's breakthroughs demonstrate that AI-RAN can deliver carrier-grade performance with continuous operation on shared NVIDIA infrastructure, enabling scalable, AI-native networks ready for commercial deployment [9]