SpaceX has a binding commitment for roughly $40 billion to finance AI chips and data center equipment, which an entity managed by Apollo Global Management, the New York-based private equity giant, will buy and lease to SpaceX, TechCrunch reported, citing a regulatory filing disclosed on Oct. 26.[1] SpaceX, which absorbed xAI in February, is using the money for GPUs, the processors that train and run AI models, Decrypt reported.[2]
The Apollo entity will buy the chips and lease them to SpaceX, which takes on about $40 billion in principal, repaid in quarterly installments over roughly five years plus interest at a fixed rate not yet set, according to Decrypt.[2] The lease comes after a tumultuous stretch in the debt markets for AI infrastructure players, TechCrunch reported, and the lease lets SpaceX fund its AI plans without relying solely on traditional debt.[1]
SpaceX expects the deal to close by year-end, subject to conditions, with leases of about five to five-and-a-quarter years starting in the first half of 2027 as equipment arrives, according to TechCrunch.[1] Apollo said the financing will help deliver Nvidia's next-generation Vera Rubin chips.[1]
SpaceX's AI segment, which includes the Grok chatbot and the Cursor coding tool, brought in $7.6 billion of the company's $14.6 billion in third-quarter revenue, while capital expenditures rose to nearly $20 billion from $6.9 billion a year earlier, TechCrunch reported.[1] SpaceX CFO Bret Johnsen said the company expects to spend $35 billion to $40 billion in the fourth quarter and is seeing "less than one-year payback periods" on its AI infrastructure investments.[1]
Elon Musk wrote on X on Oct. 25 that lenders have largely shut out companies without major revenue beyond AI data centers, predicting "guppy-level death and consolidation" among the more than 200 neoclouds, specialist firms that rent out GPU capacity.[1][2] Decrypt reported that CoreWeave, the largest neocloud, carries roughly $50 billion in debt and that its unsecured bonds yielded about 15% in mid-October.[2] Venture capitalist and entrepreneur Chamath Palihapitiya expects more AI computing power to sit idle in 2027 than forecasts assume, according to Decrypt.[2]