Advancing Spectrum Efficiency Through AI-Native Innovation
AI-RAN Alliance at the NVIDIA GTC event in Washington, D.C.
October 27–29, 2025
Booth 704, Telecommunications Pavillion
At GTC, the AI-RAN Alliance is showcasing six cutting-edge demonstrations from leading institutions and industry partners that spotlight AI’s transformative role in wireless networks. These real-world applications reveal how AI can optimize spectrum usage, enhance network intelligence, and strengthen infrastructure resilience—unlocking new revenue streams and maximizing asset utilization.
Visit us in Booth 704 and explore the demos below:
Adaptive & Privacy-Preserving Edge AI Model Partitioning via Spectrum Sensing (SUTD)
Remote Human-Robot Collaboration in XR Using AI-on-RAN Infrastructure (Northeastern University)
Experience seamless teleoperation between humans and robots through extended reality powered by AI-driven RAN intelligence. This demo showcases RoboTwin, a system where a remote operator wearing an XR headset controls a robot located across the country with real-time visual feedback and interaction. Edge AI modules running on the RAN infrastructure perform object detection and environment awareness, enabling the robot’s actions and the XR environment to stay perfectly synchronized.
AI-and-RAN Blueprint Platform (zTouch Networks & Northeastern University)
Introducing the first AI-and-RAN platform to evaluate and test AI-RAN coexistence with flexible GPU sharing and service orchestration. Developed in collaboration with RedHat and Northeastern, this demo lays the groundwork for scalable, intelligent wireless infrastructure.
AI-Native Air Interface for 6G (DeepSig)
This demo showcases an AI-Native autoencoder-based PHY that can run pilot-free on top of CP-OFDM and DFT-OFDM to improve 6G capacity, coverage, and efficiency. It highlights how OmniPHY Axon can improve throughput and PAPR on MIMO/SIMO channel simulations and ray tracing, making it a candidate for 6G Air Interface and other wireless systems.
Data4AI in VALOR (Viavi & NVIDIA)
This demo presents the baseline synthetic data generation setup for the Data4AI initiative in VIAVI’s VALOR lab, enabling robust AI training and algorithms validation for next-gen RAN applications, and ensuring spectrum-aware models can be developed without compromising privacy.
Multimodal Digital Twins for 6G RAN in Dynamic Environments (University of Oulu)
See how real-time sensor fusion and GPU-accelerated ray tracing generate channel insights directly from 3D sensor data. This demo showcases multimodal digital twins that operate without perfect 3D models in dynamic environments, which could help enable proactive, AI-driven, RAN optimization.