NVIDIA AI Aerial
NVIDIA AI Aerial is a suite of accelerated computing platforms, software, and services for designing, simulating, and operating wireless networks. NVIDIA AI Aerial contains hardened RAN software libraries for telcos, cloud service providers (CSPs), and enterprises building commercial 5G networks. Academic and industry researchers can access Aerial on cloud or on-premises setups for advanced wireless and AI and machine learning (ML) research for 6G.
Join 6G Developer Program Member Area
NVIDIA AI Aerial Building Blocks for Wireless Research and Development
Aerial CUDA-Accelerated RAN
NVIDIA Aerial CUDA-Accelerated RAN is a framework for building commercial-grade, software-defined, and cloud-native 5G and future 6G radio access networks. It includes NVIDIA GPU-accelerated interoperable PHY and MAC layer libraries that can be easily modified and seamlessly extended with AI components.
Aerial Omniverse Digital Twin
The Aerial Omniverse Digital Twin enables physically accurate simulations of complete 6G systems, from a single tower to city scale. It incorporates software-defined RAN and user-equipment simulators, along with realistic terrain and object properties. Using the Omniverse Aerial Digital Twin, researchers will be able to simulate and build base-station algorithms based on site-specific data and train models in real time to improve transmission efficiency.
Aerial AI Radio Frameworks
Aerial AI Radio Frameworks provide seamless integration with popular libraries and frameworks like PyTorch and TensorFlow, leveraging NVIDIA GPUs for generating and capturing data and training AI and ML models at scale. The frameworks include pyAerial, Aerial Data Lake, Aerial RAN CoLab Over-The Air (ARC-OTA) and NVIDIA Sionna, the leading link-level research tool for AI/ML-based wireless simulations.
Use Cases
Building Commercial 5G Networks
Accelerated 5G
NVIDIA AI Aerial enables the building of commercial-grade, high-performance, software-defined, cloud-native 5G networks, accelerated on NVIDIA infrastructure (i.e., GPU, DPU, and CPU). For the RAN L1 and L2 layers, Aerial supports GPU acceleration, delivering a high-performance virtualized distributed unit (vDU) for 5G networks.
Generative AI and 5G Data Centers
NVIDIA AI Aerial enables telcos, CSPs, and enterprises to build and deploy multi-tenant data centers capable of running both 5G and AI workloads in the cloud, at the edge, or on-premises. Examples of 5G workloads include accelerated 5G vDU and other virtualized 5G network functions like User Plane Function (UPF). These workloads run on the same data center with other orthogonal AI/generative AI workloads, such as model training (e.g., NVIDIA NeMo™) and offline image recognition (e.g., NVIDIA Metropolis). NVIDIA Aerial CUDA-Accelerated RAN is fully virtualized and cloud-ready, making it a pioneering RAN-in-the-cloud solution for the industry.
Contact the Aerial Team
Accelerating 6G Research
Cloud-Based Tools
NVIDIA 6G Research Cloud Platform provides access to the Aerial CUDA-Accelerated RAN, Aerial Omniverse Digital Twin and Aerial AI Radio Frameworks to academic and industry-based researchers. This suite of tools and frameworks enables research spanning RAN L1, L2, and core nodes. A primary focus for the 6G Research Cloud Platform is to facilitate research that brings AI/ML to all layers of the telecom stack.
The system-level simulation (i.e. Aerial Omniverse Digital Twin) and link-level simulation (i.e Sionna) setups can be accessed via a cloud API and hosted in the cloud as a fully software-based research environment. Additionally, the Aerial RAN CoLab Over-The-Air (ARC-OTA) testbed can be used as an on-premises, real-time hardware testbed for benchmarking innovations.
The 6G Research Cloud Platform can be accessed by subscribing to the 6G Developer Program, which provides developers with tools, documentation, and early access to software releases to facilitate cutting edge 6G research.
Join the 6G Developer Program
Benefits
100% Software Defined
NVIDIA Aerial CUDA-Accelerated RAN is a fully software-defined, scalable, and highly programmable 5G RAN acceleration platform for the L1 and L2+ layers on general purpose compute. This unique solution provides full flexibility to researchers for updating any part of their software.
Test by Deploying in a Digital Twin
Since NVIDIA Aerial Omniverse Digital Twins run the NVIDIA Aerial CUDA-Accelerated RAN stack over ray-traced channels, you can test the system-level performance and efficiency of your algorithms before deploying them in a live system.
Ideal for 6G Standardization
NVIDIA AI Aerial is uniquely positioned to help drive 6G standardization. New AI/ML use cases such as beam management, CSI compression, and positioning can be assessed with city scale simulation setups, realistic user mobility, and standard-compliant waveforms.
Industry Affiliations
NVIDIA works closely with industry groups and standards bodies to develop new technologies and contribute to industry publications and standards.
Join the 6G Developer Program
Developers and researchers can gain access to the 6G Research Cloud Platform—that hosts our latest software, tools, and resources—by joining the NVIDIA 6G Developer Program.