Nvidia is collaborating with T-Mobile, Cisco, and several others to develop an AI framework for 6G. Industry analysts have shared their insights on the upcoming 6G network stack. Will AI play a dominant role in 6G standards? Nvidia aims to establish an early advantage in AI for 6G. Recently reported by Fierce, Nvidia is teaming up with T-Mobile, MITRE, Cisco, ODC, and Booz Allen Hamilton to create an AI-native network stack for 6G. This initiative builds on Nvidia's AI Aerial platform and follows the unveiling of their 6G research cloud platform at the GTC conference in 2024. Despite the buzz surrounding 6G, it will be at least four and a half years before the new standard becomes commercially available. Standardization by 3GPP is not anticipated until 2028 with the release of version 21. Why is Nvidia emphasizing AI for 6G? Fierce sought insights from analysts regarding Nvidia's proactive stance ahead of the commercial standard's launch. “Nvidia is focusing on getting ahead of the curve to embed their technology into the 6G standards, ” stated Joe Madden, an analyst at Mobile Experts, in an email to Fierce. “That’s how the industry operates. ” Daryl Schoolar, an analyst at Recon Analytics, noted, “NVIDIA's collaboration with T-Mobile, Cisco, and others aims to clarify how to integrate AI into the 6G standard, addressing key questions about its specific requirements and integration into the standards. By resolving these questions, NVIDIA and its AI RAN partners hope to shape the standards' content. ” The objective is to create 6G infrastructure that leverages Nvidia’s preferred AI chips, specifically its graphical processing units (GPUs), rather than the x86 chips and custom ASICs that are currently favored by operators. “NVIDIA aims to gain traction within the research community, garner ecosystem support, and possibly attract non-incumbents—like networking vendors with mobile core capabilities but lacking mobile RAN assets, such as Cisco—to invest in the next innovation wave utilizing GPUs instead of traditional ASICs, ” explained Roy Chua, lead analyst at AvidThink. Where do the major RAN vendors stand? Nvidia’s latest collaboration for 6G does not include any leading RAN vendors.

However, Chua highlighted that the company is already engaged with Ericsson and Nokia. It may take until the next decade's arrival of 6G for the costs, complexity, and power demands of AI RAN to be effectively managed. Intel’s CTO recently remarked to Fierce that an AI RAN is not cost-effective at this stage. Nonetheless, AI-native 6G is an evolving competitive landscape. Madden expressed skepticism that Nvidia's approach would emerge as the only viable method for creating an AI-native wireless network. “Nvidia lacks substantial backing from operators and vendors within the RAN community, making it improbable that their specific AI hardware initiative will become the cornerstone for 6G, ” Madden asserted. The evolution of efforts to define and standardize AI-native elements in 6G will unfold over the next few years. It is certain that Nvidia, Intel, Qualcomm, and other key players will have significant input on the future of AI-native wireless networks for 6G.