Nvidia is in talks with Apollo Global Management to create a network of small AI data centers. The plan involves retrofitting existing buildings owned by a struggling telecom company. This strategy aims to give Nvidia a faster way to activate its chips.
The proposed network could eventually cover twenty U.S. states. Apollo would manage the construction of these sites using the telecom firm’s existing physical infrastructure. Nvidia is reportedly seeking this path to bypass the long timelines associated with building new facilities from scratch.
This move would allow Nvidia to rapidly scale its AI infrastructure across a wide geographic area. It addresses the current bottleneck of slow chip deployment times. Developers and enterprises relying on Nvidia hardware should expect a significant acceleration in infrastructure availability and regional expansion.
