NVIDIA is accelerating the development of AI-native 5G and 6G networks by open-sourcing its Aerial software, a move set to transform the telecom industry. The company will release Aerial, including CUDA-accelerated RAN, Aerial Omniverse Digital Twin, and the new Aerial Framework, on GitHub under the Apache 2.0 license starting in December, with the full digital twin tool expected in March 2026. This open-source release makes powerful wireless software accessible to a broad community of developers, researchers, and engineers, enabling rapid prototyping and faster transition from research to real-world deployment. Unlike past generations constrained by proprietary systems and limited access, the new approach allows for open, collaborative innovation at the speed of AI. NVIDIA’s Aerial software suite includes tools for building full-stack, AI-native radio access networks. These capabilities have already enabled the first U.S.-built AI-native wireless stack, demonstrating early 6G applications such as spectrum agility and integrated sensing and communications. The release is paired with the availability of NVIDIA DGX Spark, the world’s smallest AI supercomputer, now optimized for AI-native 5G and 6G research. DGX Spark delivers high performance in a compact, cost-effective form factor and can be deployed via manufacturers or through NVIDIA DGX Cloud. It supports the NVIDIA Sionna Research Kit and the Aerial Testbed, allowing teams to build, test, and validate complete wireless networks in hours—complete with user equipment, antennas, radio systems, and core networks—right from unboxing. The platform enables over-the-air testing of AI-accelerated networks using real-world data across diverse environments. Dell Technologies is also launching the Dell Pro Max with GB10, a DGX Spark-based system designed for demanding telecom research, offering a stable, high-performance environment for 6G development. Leading institutions including Northeastern University, Virginia Tech, Arizona State University, and MIT’s WINSLab and LIDS are already using NVIDIA’s AI Aerial portfolio for cutting-edge research. The AI-RAN Alliance, a global consortium of over 100 telecom leaders, is using these tools to define the architecture of future AI-native networks through shared labs, benchmarking, and live demonstrations. Alex Jinsung Choi, chairman of the AI-RAN Alliance, highlighted that open-source Aerial and DGX Spark empower developers to create modular, software-defined systems and experiment freely—from labs to live deployments—driving advances in spectrum efficiency, network performance, and new AI applications. By lowering barriers to entry and fostering global collaboration, NVIDIA is ushering in a new era of wireless innovation. This initiative strengthens U.S. leadership in open-source technology and positions the telecom industry to evolve at the pace of AI, shaping the future of connectivity and setting global standards for next-generation networks.