NVIDIA is set to release open-source software, Aerial, for building AI-native 5G and 6G networks, leading to rapid prototyping and product development in hours. This move breaks down barriers to innovation in the telecom industry, with features such as customizable pipelines and advanced channel estimation improving wireless performance.

The upcoming Aerial open-source release includes Aerial Framework for converting Python code into high-performance CUDA code, AI-powered neural models, and a dApp framework for real-time physical layer data access. This shift allows developers to experiment and build AI-native network solutions without restrictions, accelerating the transition from research to deployment with early 6G applications.

NVIDIA DGX Spark, the world’s smallest AI supercomputer, is now available for AI-native 5G and 6G research, offering the performance to run Aerial or Sionna software. Collaborators like Dell Technologies are introducing systems powered by DGX Spark for handling intensive 5G and 6G workloads, empowering researchers worldwide to tackle complex challenges in wireless technology.

Thousands of wireless innovators worldwide are leveraging NVIDIA’s AI Aerial portfolio for research and development, including leading institutions such as MIT’s WINSLab and LIDS. The AI-RAN Alliance, with over 100 telecom industry pioneers, is shaping the core architecture of AI-native wireless networks using NVIDIA AI Aerial, fueling innovations that boost spectrum efficiency and enhance network performance.

NVIDIA’s commitment to open access and global collaboration is catalyzing a wave of 5G and 6G collaboration, shaping national competitiveness and global standards. This pivotal milestone for the telecom industry enables a fully inclusive, software-defined, and AI-powered future where innovation moves at the speed of AI, marking a new chapter in wireless innovation for the AI era.

Read more at NVIDIA: NVIDIA Open Sources Aerial Software to Accelerate AI-Native 6G