NVIDIA and Global Telecom Leaders Commit to Build 6G on Open, AI-Native Platforms

NVIDIA and a coalition of global telecom operators have jointly committed to building 6G networks on open, AI-native, and secure platforms — an architectural choice that would bake AI into the network layer from the ground up rather than adding it on top of existing infrastructure. The announcement reflects a fundamental rethinking of what a wireless network is: not a pipe that carries data, but a system that makes intelligent decisions about how to route, prioritize, and manage that data in real time.

The distinction matters. Today's 5G networks use AI as an optimization tool layered on top of conventional network architecture — adjusting tower power, predicting congestion, reducing interference. In an AI-native network, those capabilities aren't bolted on. They're the architecture. The network itself is trained, updated, and operated through AI models, with open interfaces that allow operators to customize and extend the intelligence layer.

Why Telecom Operators Are Committing Now

An NVIDIA survey of telecom operators included in the announcement found that operators now view AI as a strategic tool for network reliability and grid load management — not primarily a cost-reduction play. That's a meaningful shift from two years ago, when most carrier AI conversations centered on reducing headcount in network operations centers (NOCs) and automating tier-one customer service.

The reliability framing connects directly to the AI infrastructure buildout. As hyperscale data centers demand more power and more bandwidth with stricter latency requirements, the carriers serving those facilities face SLAs — service level agreements — that older network architectures can't consistently meet. AI-native network management offers a path to the kind of dynamic, self-healing network behavior that enterprise AI buyers increasingly require.

What "Open" Means and Why It's Contested

The "open" in open AI-native 6G refers to the Open RAN (Radio Access Network) framework, which disaggregates the software and hardware components of a base station so that operators can mix vendors rather than being locked into a single provider's proprietary stack. Ericsson, Nokia, and Huawei have historically dominated the integrated base station market. Open RAN shifts market power toward software vendors — including NVIDIA, whose GPUs increasingly power the baseband processing that used to be done by custom silicon in proprietary hardware.

NVIDIA's motivation in pushing AI-native open 6G is straightforward: if AI processing moves to the network layer and runs on general-purpose accelerated computing rather than proprietary hardware, the addressable market for NVIDIA silicon expands significantly. The company's involvement in this coalition isn't philanthropy; it's a positioning move in a multi-trillion dollar infrastructure transition.

The tension is that major incumbent vendors — particularly Ericsson and Nokia — have invested heavily in their own AI capabilities on top of their proprietary systems. The open coalition's success depends on operators being willing to take on integration complexity in exchange for vendor flexibility.

Where 6G Stands in the Deployment Timeline

6G remains a standards and research phase technology. The 3GPP, the international standards body for mobile networks, has 6G work items beginning in earnest in 2025, with early specifications expected around 2028 and commercial deployments possible in the early 2030s. The U.S., EU, South Korea, Japan, and China are all running parallel national 6G programs with significant government funding.

Commitments made now shape what those standards look like. When major telecom operators and a company of NVIDIA's scale announce joint platform commitments before standards are finalized, they're participating in the standards process — pushing their preferred architecture toward the default.

Key performance targets for 6G under current research goals:

• Peak speeds: Up to 1 terabit per second (roughly 100x faster than 5G peak)
• Latency: Sub-millisecond in target configurations
• Connection density: 10 million devices per square kilometer (10x 5G)
• Energy efficiency: 10-100x improvement per bit transmitted

What to Watch

The standards process is the critical near-term watchpoint. Look for 3GPP Release 20 and 21 milestones, which will begin to formalize 6G architecture choices. Also watch for spectrum policy decisions — particularly in the upper millimeter wave bands and terahertz frequencies being evaluated for 6G capacity. In the U.S., FCC spectrum auctions for 6G-relevant bands are expected in the late 2020s. Who participates and at what prices will signal how aggressively U.S. carriers are moving.

Source: NVIDIA Newsroom — [NVIDIA and](/manufacturing/nvidia-hannover-messe-physical-ai-factories) Global [Telecom Leaders](/telecom/india-coai-digicom-summit-ai-5g-networks) Commit to Build 6G on Open and Secure AI-Native Platforms