SpaceX Just Locked In $40 Billion for AI Chips. The Compute Arms Race Is Escalating.

The AI race is increasingly being decided by something less glamorous than model demos: access to chips, power and capital. SpaceX has now put a number on just how expensive that race is becoming.

The company disclosed a financing arrangement with an affiliate of Apollo Global Management involving roughly $40 billion in principal payments. Apollo will acquire AI chips and server hardware and lease the equipment to SpaceX, with the leases expected to begin after hardware deliveries in 2027.

Why the deal matters

This is not a conventional equipment purchase. It is effectively a large-scale infrastructure financing structure built around scarce computing hardware. Each lease is expected to run a little more than five years, giving SpaceX access to enormous amounts of compute without paying the entire hardware bill upfront.

That matters because frontier AI is becoming a balance-sheet game. The companies that can secure GPUs, power and data-center capacity at scale have an advantage that smaller competitors may struggle to match.

AI and space are becoming part of the same strategy

SpaceX is spending aggressively across Starship, satellites and AI infrastructure. The company has said it expects tens of billions of dollars in capital spending, and its AI ambitions are increasingly tied to the broader economics of Starlink and future space operations.

The bigger takeaway is that AI infrastructure is beginning to look more like energy, telecom and aerospace infrastructure: long-lived, capital-heavy and financed over years rather than quarters.

What to watch next

The next question is how quickly SpaceX can put the new hardware to work and whether its AI businesses can generate enough revenue to justify this scale of investment. The chip deliveries are expected to arrive in 2027, which means the spending story is only beginning.

For the broader industry, the deal is another sign that the AI boom is moving beyond software. The next stage may be defined by who can finance and operate the physical systems underneath it.