NVIDIA is moving beyond data center GPUs into telecom infrastructure, betting that 6G base stations will become the next frontier for AI compute distribution.
NVIDIA is moving beyond data center GPUs into telecom infrastructure, betting that 6G base stations will become the next frontier for AI compute distribution.

NVIDIA is moving beyond data center GPUs into telecom infrastructure, betting that 6G base stations will become the next frontier for AI compute distribution.
NVIDIA is pushing into telecom infrastructure, seeking Chinese base station manufacturers to build 6G AI-RAN systems that combine wireless connectivity with AI compute — a market Omdia projects will exceed $200 billion cumulatively by 2030. The company is accelerating its entry into the telecom operator market and searching for partners in China to develop 6G base stations meeting overseas requirements, according to a core insider in the communications industry chain who spoke to Jiemian News.
"NVIDIA is accelerating its entry into the telecom operator market and is seeking base station manufacturer partners in China to develop 6G base stations that meet overseas market requirements," the insider said, describing the project's pace as "quite urgent."
Shenzhen Jiaxian Communication is one of the partners NVIDIA is building relationships with. NVIDIA first contacted the company in 2025, and the two have been collaborating for more than six months on 6G AI-RAN base stations built on NVIDIA's CUDA ecosystem. A working group of more than 10 NVIDIA staff across Singapore, Hong Kong and mainland China is studying the CUDA open-source ecosystem, according to a Jiaxian insider. The companies have not ruled out establishing a commercial entity for joint overseas expansion.
Jiaxian plans to release a 6G AI-RAN prototype based on NVIDIA's CUDA ecosystem by the end of 2026. If the product reaches scale deployment, the company projects sales of one million units across the 6G era, representing roughly 200 billion yuan ($28 billion) at approximately 200,000 yuan per station. NVIDIA does not manufacture complete base stations — it supplies GPUs and the CUDA software stack, while partners handle RF, microwave, antenna, AAU, manufacturing and operator network integration.
Why China's base station supply chain matters
China accounts for 60-70 percent of global 5G base station hardware manufacturing, according to research firm IndexBox, with production clusters in Shenzhen, Dongguan, Shanghai and Chengdu. Even Nokia and Ericsson — NVIDIA's telecom partners — manufacture in China through contract manufacturers including Luxshare Precision and Foxconn.
"Even Nokia and Ericsson have their production capacity in China," the industry insider said. "The base station industry has essentially disappeared overseas — it's deeply rooted in China and hard to replicate elsewhere."
NVIDIA's telecom push extends beyond the Chinese partnership. In October 2025, NVIDIA invested $1 billion in Nokia for approximately 2.9 percent of the company, with Nokia agreeing to develop commercial AI-RAN products on NVIDIA's platform. In February, NVIDIA joined Nokia, Ericsson, T-Mobile, Deutsche Telekom and SoftBank in proposing an open, software-defined, AI-native 6G platform. In March, NVIDIA, T-Mobile and Nokia announced AI-RAN test progress, running edge AI tasks on base stations alongside 5G connectivity.
Operators shift to token-based pricing
Chinese operators are preparing for the AI base station era. China Telecom launched Token packages in May 2026, and its Ningxia subsidiary's "Token factory" procurement reached approximately 17.4 billion yuan. China Mobile is experimenting with bundling communication traffic with large-model tokens. Industry insiders estimate 6G base stations will trigger a new round of trillion-yuan telecom infrastructure investment.
NVIDIA shares closed at $219.22 on Aug. 5, up 3.43 percent, after the stock reached a new high since June 4 at $220.43. The AI-RAN expansion gives NVIDIA a path into edge computing as data center GPU demand matures. Qualcomm CEO Cristiano Amon captured the stakes in January: "Whoever wins edge AI is more likely to win the entire AI era."
This article is for informational purposes only and does not constitute investment advice.