Executives from T-Mobile US and Nvidia touted the promise of AI RAN during a Nokia online summit held this week, which included discussions on the impact the technology would have across networks, applications and customer service.

Speakers from the two companies, in addition to representatives from SoftBank Corp and Nokia took part in the vendor’s Tech Winter Horizon event. All of the companies are among the members of the AI-RAN Alliance announced at MWC Barcelona 24.

Nokia executives explained AI RAN is parsed across three areas in the AI-RAN Alliance: AI-for-RAN, AI-and-RAN and AI-on-RAN.

AI-for-RAN improves traditional building blocks of RAN by implementing AI to drive spectral efficiencies while AI-and-RAN optimises infrastructure and unlocks monetisation opportunities.

Steph Delvoye, head of marketing at Nokia mobile networks, stated AI-on-RAN is using AI to augment the RAN itself for services at the edge.

“When you combine AI with RAN, you create an engine for future evolution. AI enables RAN innovation and RAN enables AI innovation,” he said.

Nvidia
Ronnie Vasishta, SVP of telecom at Nvidia, stated the current RAN infrastructure is single use only. He noted the capex spent to support RAN and is on average only 30 per cent utilised.

“Embedding AI as well as RAN as software-defined workloads onto the accelerated compute infrastructure has opened a wealth of new business opportunities,” Vasishta explained.

Those opportunities include hosting AI applications closer to the point of use.

“Applications such as video search, robot control, factory automation, and servicing generative AI digital humans,” he said.

Vasishta noted an operator recently calculated that deploying one AI RAN server at 6,000 base stations will yield $1 billion per year through the generation of AI tokens.

“Our collaboration with Nokia and T-Mobile will help unlock the full potential of AI RAN and drive the future of telecommunications.”

T-Mobile
Karri Kuoppamaki, SVP of technology development and strategy at T-Mobile (pictured), said AI RAN is fundamental to building multi-purpose networks that improve customer experience, increase network efficiencies and add capacity.

“AI with RAN is about maximising the capital efficiencies of our network by being able to do both AI workloads and telco workloads at the same time,” he noted. “If we can realise this vision, it has the potential of changing how wireless networks are built in the future where AI will revolutionise wireless, and wireless will revolutionise AI. And if that’s not exciting, I don’t know what is.”

SoftBank
SoftBank’s Ryuji Wakikawa, VP, head of the research institute of advanced technology, said in addition to AI RAN being key for creating new revenue streams, 6G centimetre spectrum and AI will enhance performance, connectivity and bandwidth.

“AI will be deeply embedded in the 6G network and algorithms,” added Ari Kynaslahti, CTO and SVP of mobile network strategy and technology at Nokia. “In 6G, AI will be the foundation for AI interface design and optimisation.”