SpaceX Secures $40 Billion to Expand AI Infrastructure
SpaceX is investing in 10 gigawatts of AI compute capacity by 2027, supported by a $40 billion financing deal with Apollo Global Management.
Elon Musk announced that SpaceX will aggressively expand its artificial intelligence infrastructure, targeting at least 10 gigawatts of compute capacity by 2027. During a third-quarter earnings call on October 20, Musk set a goal for annual recurring revenue between $500 billion and $1 trillion by the end of 2027. To support this growth, SpaceX has committed to using Nvidia chips exclusively, standardizing on the Vera Rubin architecture and collaborating with the semiconductor firm to develop radiation-resistant chips for orbital data centers.
To finance the acquisition of this hardware, SpaceX entered a $40 billion deal with an affiliate of Apollo Global Management. Under the agreement, Apollo will purchase the AI chips and server hardware and lease them back to SpaceX for terms slightly exceeding five years. Musk announced the closure of this debt financing on Sunday, noting that the funds primarily target equipment arriving in the first and second quarters of 2027.
SpaceX intends to use the AI computing expansion to help fund its long-term missions to the Moon and Mars. While the Apollo deal will be finalized by the end of 2026, CFO Bret Johnsen indicated that the company still expects capital expenditures between $35 billion and $40 billion for the fourth quarter of 2026.
