Telecom operators are innovating beyond voice and data services to optimize their wireless networks with AI computing infrastructure and meet next-generation AI requirements such as mobile, robots, self-driving cars, smart factories, and 5G.

NVIDIA today unveiled NVIDIA AI Aerial, which optimizes wireless networks and delivers next-generation AI experiences on a single platform.

The recently launched NVIDIA AI Aerial is a family of accelerated computing software and hardware for designing, simulating, training, and deploying AI radio access networks (AI-RAN), the wireless networks of the AI era.

As a critical foundation for large-scale network optimization to meet the demands of a variety of new services, it enables telecommunications operators to reduce total cost of ownership and create new revenue opportunities for enterprise and consumer services.

NVIDIA AI Aerial helps telecom service providers enable remote work for manufacturing robots and self-driving cars. △It enables computer vision in manufacturing, agriculture, and logistics, △Generative AI-based copilots and personal assistants, △New spatial computing applications, △Robotic surgery, △3D collaboration, and △5G and 6G advancements.

Key partners in the NVIDIA AI Aerial ecosystem include SoftBank and Fujitsu, while Ansys and Keysight are involved in test and simulation systems. Partners such as DeepSig, ETH Zurich, Northeastern University, and Samsung are collaborating in academia on 6G research and the NVIDIA Aerial AI Radio framework.

In addition, the ecosystem is formed by Arna Networks, Canonical, Red Hat, Wind River, Arcus Networks, Dell Technologies, Hewlett Packard Enterprise, Supermicro, and Vapor.io.

■ AI-RAN Leading Next-Generation Networks

AI-RAN is the foundation of future multi-purpose networks that leverage AI-based communications capabilities, and NVIDIA AI Aerial is the world’s first AI-RAN platform that can host generative AI and RAN traffic, as well as integrate AI into network optimization.

AI-RAN offers new revenue opportunities through high-performance, energy-efficient software-defined RAN, enhanced network experience, and edge AI applications for hosting internal or third-party generated AI applications.

■ Utilizing NVIDIA AI Aerial for the Telecommunications Industry

The NVIDIA AI Aerial Platform provides access to the full range of capabilities, including training, simulation, and inference, along with high-performance software-defined RAN. This will enable telecommunications operators to participate in every stage of development through deployment for next-generation wireless networks.

NVIDIA Aerial CUDA Accelerated RAN includes software libraries that enable partners to develop and deploy high-performance virtualized RAN workloads on NVIDIA accelerated computing platforms.

The NVIDIA Aerial AI Radio framework includes PyTorch and TensorFlow-based software libraries for developing and training models to improve spectral efficiency and add new capabilities to 5G and 6G wireless signal processing. It includes NVIDIA Sionna, a link-level simulator that provides development and training of neural network-based 5G and 6G wireless algorithms.

NVIDIA Aerial Omniverse Digital Twin (AODT) is a system-level network digital twin development platform. AODT enables physically accurate simulation of wireless systems, from single base stations to comprehensive networks with multiple base stations spanning an entire city.

AODT integrates software-defined RAN (Aerial CUDA-accelerated RAN) and user equipment simulators with realistic terrain and object properties of the real world.

■ NVIDIA AI Aerial and AI-RAN Innovation Center

NVIDIA is collaborating with T-Mobile, Ericsson, and Nokia to establish the AI-RAN Innovation Center and accelerate the commercialization of AI-RAN.

The center will leverage key capabilities of the NVIDIA AI Aerial platform. This collaboration focuses on further combining RAN and AI innovations to deliver innovative network experiences to customers through AI-RAN development.

“AI-RAN will transform the telecom industry, and the establishment of the AI-RAN Innovation Center will help accelerate industry collaboration on this journey,” said Tommi Uitto, President, Mobile Networks, Nokia. “By bringing together leading companies in telecom and AI, we can unleash the potential of AI in the network to improve performance, reduce costs and create new opportunities for our customers.”

“Ericsson has been investing in AI-RAN solutions to enable communication service providers to deploy portable RAN software running on multiple platforms,” said Fredrik Jejdling, SVP and head of Ericsson Networks. “We are currently evaluating the performance and cost of NVIDIA accelerated computing in this context.”