SpaceX has secured a binding commitment for roughly $40 billion in financing to acquire and lease artificial-intelligence computing hardware, including Nvidia’s next-generation Vera Rubin GPUs. The financing is being arranged through an entity managed by Apollo Global Management, with Goldman Sachs acting as structuring agent and additional commitments from funds linked to BlackRock, KKR and Goldman Sachs Asset Management. SpaceX called the transaction the largest committed GPU financing ever.
How the deal works
Under the structure, the financing vehicle will acquire GPUs and related data-center equipment and lease the hardware to SpaceX.
The master finance lease is expected to close before the end of 2026, subject to closing conditions. Individual lease schedules are expected to begin when equipment is delivered in the first half of 2027.
Why SpaceX needs GPUs
SpaceX is no longer only a rocket and satellite company. It is building large terrestrial computing capacity to support AI workloads and cloud services.
The company said the infrastructure will help power Grok for enterprise and consumer users as well as future cloud customers.
Nvidia Vera Rubin is central to the plan
The financing specifically includes Nvidia’s Vera Rubin generation of AI chips. These processors are designed for extremely large training and inference workloads.
Access to leading-edge GPUs has become one of the most important competitive advantages in AI because supply remains expensive and capacity is constrained.
Leasing reduces upfront capital pressure
Instead of paying the entire hardware cost immediately, SpaceX can spread payments across multi-year leases. Reuters reported that individual leases are expected to run slightly longer than five years from delivery.
This structure can preserve cash for other capital-intensive projects while still allowing the company to deploy computing infrastructure quickly.
SpaceX already spends heavily
SpaceX is simultaneously funding launch vehicles, Starlink satellites, ground infrastructure and other projects. Management has indicated that capital spending in the final quarter of 2026 alone could reach tens of billions of dollars.
Adding a massive AI buildout increases the need for diversified financing sources.
Apollo’s role reflects private-credit growth
The deal also demonstrates how alternative asset managers are becoming major financiers of AI infrastructure. Traditional bank loans and corporate bonds are no longer the only sources of capital for data-center expansion.
Private-credit firms can structure leases and asset-backed financing around expensive hardware with relatively predictable depreciation and demand.
The broader AI infrastructure boom
AI companies and technology groups are racing to secure chips, power and data-center space. Meta is planning a major equity raise, Microsoft has forecast enormous capital spending and other companies are using debt and project finance to expand capacity.
SpaceX’s $40 billion transaction places it among the largest participants in that infrastructure race.
Questions about utilization
Buying or leasing GPUs at this scale only makes sense if the hardware remains highly utilized. SpaceX says its computing investments are producing short payback periods, but investors will watch whether demand remains strong enough to justify the financing.
Rapid changes in chip technology also create obsolescence risk. Hardware financed for more than five years could be surpassed by newer systems well before lease payments end.
Why SpaceX may have an advantage
The company controls a large satellite network, launch infrastructure and a growing communications platform. That ecosystem could create internal demand for AI and cloud services while also supporting external customers.
Its ability to launch satellites at scale may also give it options for integrating terrestrial and space-based computing in ways that conventional cloud companies cannot easily replicate.
The financing race is becoming strategic
The AI industry increasingly depends not just on software talent but on the ability to finance and deploy physical infrastructure. GPUs, power generation, cooling systems and data centers require enormous upfront investment.
SpaceX’s Apollo-led transaction shows how technology companies are turning to complex financing structures to secure capacity before competitors do. If the deal closes as planned, deliveries beginning in 2027 could significantly expand SpaceX’s computing footprint and deepen its role in the global AI infrastructure market.
The financing also gives SpaceX more predictability over hardware procurement at a time when advanced AI chips are in exceptional demand. Locking in a large pool of Vera Rubin systems early could reduce the risk of shortages delaying data-center expansion. The tradeoff is a long-term payment commitment tied to equipment whose economic value will depend on utilization and the pace of future chip improvements.
