NVIDIA is making significant strides into the telecommunications operator market, with plans to team up with Chinese base station manufacturers to co-develop 6G AI-RAN base stations. These are expected to enter the trial network phase by 2027 or 2028. Shenzhen Jiaxian Communication has emerged as a pivotal partner in this endeavor. The two companies will leverage the CUDA ecosystem for their research and development efforts, aiming to unveil a prototype by the end of this year and introduce it to the market within the next one to two years. It is anticipated that, should large-scale deployment be realized, Jiaxian Communication's base station sales in the 6G era could soar to the million-unit mark.

In addition to its collaboration with Jiaxian Communication, NVIDIA is also working alongside companies such as Nokia to collectively propel the advancement of AI-RAN and 6G infrastructure. Presently, the AI infrastructure landscape is undergoing a transformation, shifting from centralized computing centers to distributed computing networks. This shift is driven by a burgeoning overseas demand for terminal edge computing capabilities.

NVIDIA's decision to partner with Chinese suppliers is primarily motivated by the fact that the global base station industry chain is heavily concentrated in China. Base station R&D and production are highly dependent on China's industrial expertise, talent pool, and manufacturing capacity. Moreover, there are few alternatives for independent Chinese base station partners. Concurrently, operators are also expediting the adoption of AI base stations, seeking to reinvent themselves by offering model invocation and inference services. The deployment of 6G base stations is poised to unleash trillion-yuan-level investments in communication infrastructure, with rapid advancements being made in end-side and distributed computing infrastructure.