It might be argued that Nvidia has made 6G à la mode.

Nvidia’s most recent GTC event saw the chip giant debut Aerial RAN Computer Pro (ARC-Pro), a 6G-native accelerated computing platform aimed at enabling telecom operators to shift from 5G-Advanced to 6G.

The announcement came alongside Nvidia's $1 billion equity investment in Nokia, helping the Finnish vendor’s development of 5G and 6G radio access network (RAN) technologies – and arguably adding to its street cred for a hot minute or two in the eyes of tech bros.

This might be a welcome reprieve from the usual skepticism concerning 6G. As reported by SDxCentral, telecom operators are not necessarily rushing to deploy 6G technology, with many still absorbing major 5G network costs, and seeing little return on investment.

According to Alain Mourad, senior director and head of wireless Lab Europe at InterDigital, that skepticism keeps him “grounded” as he leads the wireless R&D firm’s future-facing development of 6G technologies.

The U.K.-based Mourad also works on 6G at ETSI (European Telecommunications Standards Institute), specifically as chair of the industry specification detail for integrated sensing and communication (ISAC) work in 6G.

That development primarily revolves around mapping where and how radar-like ISAC will be used, as well as channel modeling for ISAC to see how radio channel models change once they sense objects which are not part of a network.

ETSI’s ISG ISAC group also identifies system and RAN architecture capable of support ISAC capabilities, as well as exploring the implications of sensing technology from the standpoints of security, privacy, and trustworthiness.

Mourad talks about security mainly in the context of 6G systems “seeing” objects even when they’re not connected to the network (and not, as a certain U.S president might suggest, giving a deeper view into someone's skin).

“There is a really big deal of privacy here … so how do we handle the privacy for these objects?” said Mourad, flagging the risk that the radio signals used for sensing could be exploited by unintended parties. As such, ETSI is examining how to protect the sensing signal from such misuse by managing access and exposure on the network.

He also highlighted “trustworthiness” as a third dimension alongside security and privacy, as users in a 6G future need to be able to assess how valid and reliable the sensing data is, and under what assurances it can be safely used.

For Rajesh Pankaj, the U.S.-based chief technology officer at InterDigital, the ISAC aspect is a key pillar of 6G, built into the way radios and infrastructure operate and create a network that “acts as a big, huge sensor.”

Existing radio infrastructure will be used not just for connectivity, he believes, but as a massive distributed sensor, with radio signals helping future neo-networks map their surroundings, detect objects, and infer occupancy or movement.

This capability could support a range of applications, from security and room occupancy detection to yet-unimagined services, with the CTO pointing to how the possibilities of positioning and connectivity were not first obvious when the underlying technologies were first standardized.

“You use Uber now without even thinking about it, and don't think about what is required for that to be done: better position, location rates, communication capability, low latency, all of that is needed before you can use something like that.”

“These apps, these capabilities can lead to more innovation and the type of applications you can run, but the main goal is that you already have the radio transmitter and receiver, both on your phone and the network, and using the phone as a sensor, you can actually map your surroundings.”

AI and the network as a sensor

Perhaps unsurprisingly, AI will be another essential 6G pillar. Mourad sees the tech running at the service, transport, and run layers of the neo-network, as well as driving the ISAC aspect and vice versa.

“It’s another degree of autonomy. Eventually, you’ll translate an intent using AI and agentic AI. Agents are making an appearance gradually, I would say, into the downstream, but it will take some time before we see these actually happening in the market.”

Ultimately, the market relies on standards, and in Pankaj’s view, the current 2030 timetable for 6G standardization is a realistic date to work to, especially considering the spectrum real estate (or lack of it). And that’s before you consider chip development, device integration, and network upgrades.

Even with business enthusiasm perhaps stoked by the Nokia/Nvidia deal, any earlier 6G deployment would ultimately compromise safety, Pankaj noted.

The next-decade timeline of 6G also coincides with the vision for self-healing, level 4 (L4) autonomous networks, with the two concepts intertwined thanks to 6G’s AI-native nature.

“The buzzword is agentic AI, but [it’s the same as] an algorithm that's running somewhere that's supposed to take care of something, and it needs to make sure that it has access to the data so it can make modifications to the way it operates,” Pankaj explained. “So now it’s just a question of whether you can take advantage of the real-time data to modify things in real time and make the network adapt to the situation … Native AI can have an impact on the interface, and that’s exactly what you need to send on the air interface."

The CTO reminded SDxCentral that slow-and-steady standards work exists to essentially define the language between a phone and a base station to communicate with each other. Without that, the network won’t work, hence InterDigital’s key work with 3GPP in defining 6G standards.

Going back to the Nokia and Nvidia tie-up, it’s worth putting 6G to one side with a reminder that AI defines the Jensen Huang giant more than radio.

Mourad sees the market play as part of a broader convergence of AI, networks, and 6G. Nvidia is likely keen on Nokia’s telco infrastructure to handle the “10-times increase in compute” brought about by AI.

Pankaj pointed to the changing nature of traffic patterns driven by AI applications themselves, shifting the downlink-heavy status quo and placing new strain and latency requirements on uplink channels.

“So in addition to making sure that AI is part of the network and improves it, we also have to have the network deal with the traffic that is created by the AI applications that everybody is using,” said the CTO.

6G, in his view, must be engineered from the outset to handle the bandwidth blitz this will bring about, something Nvidia is already working on with its 6G AI-Wireless Intelligent Network (AI-WIN) initiative, which has seen a multimodal ISAC application developed with Booz Allen that fuses camera vision and radio-frequency sensing to allow for detection and tracking of objects.

Thus, while the Nokia mega-deal may not be limited to 6G, Mourad notes it’s impossible to “decouple” Nvidia’s ambitions from 6G.

“At the end of the day, to do all this you will definitely need 6G. That’s why I say take with a bit of caution, what's kind of being promoted as decoupling all this from 6G … AI is going to happen, this announcement proves that, and also shows the industry is moving from a commercial standpoint, to give a push for even more AI – but in a reasonable way into the network.”