Elon Musk said SpaceX has closed $40 billion in debt financing to buy graphics processing units, a deal that shows just how large — and how expensive — the race to secure AI computing power has become. The announcement, made by Musk in a post on X on October 25 and reported by Reuters on October 26, follows earlier reporting that SpaceX was in talks with banks and asset managers to raise the same sum for Nvidia AI chips.
Key Takeaways
- Musk said SpaceX closed $40 billion in debt financing for GPU purchases, describing it as primarily callable debt.
- The GPUs are expected to arrive in the first and second quarters, with reports describing the deliveries as falling in the first half of 2027.
- SpaceX pointed to its investment-grade status and revenue from Starlink and launches as support for its access to credit.
Musk described the borrowing in plain terms. “This is primarily callable debt for GPU purchases that will arrive in Q1 and Q2. Not a major factor this quarter,” he wrote. Callable debt can be repaid early by the borrower, which gives a company flexibility if conditions change — but it does not make the scale of the commitment any smaller. Forty billion dollars for processors alone is, by any measure, one of the largest single AI infrastructure financings reported so far.
Why the Deal Matters
The funding underscores how the AI boom has turned access to advanced processors into a contest over capital itself. Companies building AI systems do not just need good models; they need the physical chips to train and run them, the data centres to house them, and the power to keep them running. When the best chips are scarce, the companies that can raise money fastest and cheapest gain an advantage that is hard for rivals to close.
Reports describing the financing said it was structured through a special purpose vehicle, a separate entity that keeps the debt off SpaceX’s balance sheet and helps preserve its investment-grade credit rating. Musk himself pointed to that rating, along with revenue from Starlink satellite internet and the launch business, as reasons SpaceX could still borrow at scale while other AI companies face tighter credit. That distinction — between companies with revenue outside AI data centres and companies without it — is at the heart of his warning about the wider market.
The funding underscores how the AI boom has turned access to advanced processors into a contest over capital itself.
A Warning About the Neocloud Shakeout
Alongside the financing announcement, Musk warned that debt markets have almost entirely shut off access to money for companies that lack substantial revenue independent of AI data centres. Reports of his remarks said he predicted that more than 200 so-called neocloud companies — specialist firms that rent out GPU capacity — face widespread failures and consolidation, with only a small fraction expected to survive.
That warning is a forecast, not an outcome, and it should be read as one. But it fits a pattern visible across the AI industry this year: huge capital spending by the largest players, rising concern about how that spending is financed, and growing scrutiny of business models that depend entirely on renting out chips. The debate is no longer only about which model is smartest; it is about which balance sheet can survive a downturn in the chip-rental market.
The same questions about AI’s reach are now reaching policymakers. In the United States, new rules are already emerging for AI-generated content, as covered in our report on Pennsylvania’s law requiring labels on AI-generated political and commercial ads. Financing, regulation and technology are moving together, and a $40 billion debt deal sits at the intersection of all three.
- SpaceX closed $40 billion in debt financing for GPU purchases, according to Musk.
- The debt is primarily callable, and the GPUs are due to arrive in the first and second quarters.
- Earlier Reuters reporting described talks over $40 billion for Nvidia AI chips.
- SpaceX revenue rose 161 percent to $14.58 billion in the three months ended September 30.
- Capital spending in the same quarter nearly tripled to $19.6 billion.
- Musk warned that more than 200 neocloud firms face failures and consolidation.
What SpaceX’s Own Numbers Show
The financing did not arrive in a vacuum. SpaceX’s revenue climbed 161 percent to $14.58 billion for the three months ended September 30, compared with a year earlier, according to Reuters reporting of the company’s results. Capital spending in the same quarter nearly tripled to $19.6 billion, reflecting heavy investment in AI infrastructure. The company’s chief financial officer was reported to expect $35 billion to $40 billion in capital spending in the fourth quarter of 2026 alone.
SpaceX is also looking further ahead than bought-in chips. Terafab, its chip-manufacturing joint venture with Tesla in Austin, is expected to produce its first test chip around May or June of 2027, with volume production targeted for the third quarter of 2028, according to the same reporting. Before its initial public offering, SpaceX had already listed manufacturing its own GPUs among the substantial capital expenditures it planned. Buying Nvidia chips now and building its own chips later are two halves of the same strategy: secure supply today, reduce dependence tomorrow.
For markets, the deal is another reminder of how quickly AI spending can move sentiment. Rising energy costs and geopolitical tension have already tested broader investor nerves in recent weeks, and a $40 billion borrowing by one of the world’s most closely watched private companies adds a new chapter to the question investors keep asking: who pays for the AI buildout, and what happens if the returns arrive more slowly than the bills?
What to Watch Next
What to watch next is what the article has already set out: whether the GPUs arrive in the first and second quarters as described, whether SpaceX’s capital spending stays near the levels reported for the fourth quarter, whether Terafab hits its test-chip timetable in mid-2027, and whether the neocloud shakeout Musk forecast begins to appear in failures and mergers. Until those markers arrive, the verified picture is the financing itself — $40 billion in debt, closed according to Musk, for the chips at the centre of the AI race.