Key takeaways SpaceX secures $40 billion commitment for Nvidia AI hardware Apollo and Goldman Sachs lead landmark GPU financing arrangement Five-year leases support Musk's expanding AI comput
Elon Musk's SpaceX announced a $40 billion AI hardware financing agreement on October 26, bringing Apollo Global Management, Goldman Sachs and other major Wall Street investors into one of the largest computing infrastructure deals ever disclosed.
Under the official agreement, an Apollo-managed entity will purchase Nvidia GPUs and related equipment before leasing them to SpaceX. The arrangement will finance next-generation Vera Rubin chips for the company's growing terrestrial computing infrastructure.
The agreement confirms earlier financing negotiations, although the master lease remains subject to closing conditions. SpaceX expects to finalize it before the end of 2026, with equipment deliveries and lease payments beginning in the first half of 2027.
How SpaceX's $40 Billion AI Financing Actually WorksRather than purchasing $40 billion of Nvidia equipment entirely with its own cash, SpaceX will lease hardware acquired by the Apollo-managed financing entity.
Each equipment tranche will carry a lease lasting approximately five to five-and-a-quarter years. Quarterly principal payments will generally equal one-twentieth of the financed equipment cost, with additional interest and expenses. The interest rate has not yet been determined.
Apollo and Goldman Sachs are leading the financing, with Goldman acting as sole structuring agent. Funds managed by BlackRock and Goldman Sachs Asset Management, alongside KKR-advised capital accounts, have also made anchor commitments.
The transaction illustrates how AI infrastructure borrowing is evolving beyond conventional corporate loans and bonds.
It also gives SpaceX access to additional hardware without requiring the entire purchase amount upfront, although the company still assumes substantial multiyear payment obligations.
SpaceX's AI Business Is Becoming Bigger Than StarlinkSpaceX is no longer primarily a rocket-launch and satellite-connectivity business.
Its Super Intelligence division generated $7.56 billion in third-quarter revenue, surpassing the $5.25 billion reported by its connectivity division, which includes Starlink. That AI revenue expansion helps explain the company's appetite for additional computing capacity.
Management expects fourth-quarter capital expenditures of $35 billion-$40 billion as investment in infrastructure accelerates.
The company is also selling computing capacity to outside customers. An earlier Reflection AI agreement demonstrated how its Colossus facilities could generate revenue beyond supporting Grok models.
The new hardware will use Nvidia's Vera Rubin architecture, extending a technology relationship already central to SpaceX's computing strategy.
SpaceX executives claim some infrastructure investments can pay for themselves in less than one year. However, such expectations depend on utilization, operating costs and customers continuing to purchase computing capacity.
Why Wall Street Is Financing SpaceX While AI Rivals StruggleThe timing is particularly significant because financing conditions have deteriorated for specialized AI cloud providers.
Musk recently warned that more than 200 neocloud companies could face widespread consolidation or failure as lenders become increasingly reluctant to finance businesses dependent almost entirely on AI infrastructure revenue.
Unlike many smaller cloud providers, SpaceX benefits from established rocket-launch and Starlink operations. Musk argues that this diversified revenue supports its access to investment-grade financing.
That advantage contrasts with the pressure facing CoreWeave and Nebius, whose borrowing costs and stock valuations have come under scrutiny.
