SpaceX GPU debt of $40 billion has closed, Elon Musk said on October 25, 2026, while warning that debt markets have almost entirely shut off access to money for firms without revenue independent of AI data centers. In posts on X, he said the financing is primarily callable debt for graphics processing units expected in the first and second quarters of 2027. He predicted that more than 200 neoclouds would face widespread failures and consolidation.
Key Takeaways
- SpaceX closed $40 billion in primarily callable GPU debt.
- Chips are expected in the first and second quarters of 2027.
- Musk cited an investment-grade rating plus Starlink and launch revenue.
- He predicted 200-plus neoclouds would face failures and consolidation.
What the SpaceX GPU Debt Close Includes
CNBC reported that SpaceX has closed $40 billion in debt financing to buy GPUs used to power artificial intelligence systems. A regulatory filing on Monday said the company signed a binding commitment with an Apollo-managed entity for roughly $40 billion in GPU financing. SiliconANGLE reported that Musk said the chips will come from Nvidia and other suppliers to fill SpaceX AI data centers.
Musk said he expects most of the debt to be repaid early, and that it will not have a significant impact on fourth-quarter earnings.
SiliconANGLE said the posts appear to confirm an October 7 Financial Times report that SpaceX was in talks to secure $40 billion. The Times cited Apollo Global Management as lead investor, with around $10 billion in bank loans and $30 billion in investment-grade debt, and mentioned Pacific Investment Management Company LLC among institutions providing funds. CNBC’s account of the filing confirms the Apollo-managed commitment. It does not describe that loan-and-bond split as closed terms.
Starlink Revenue and the Investment-Grade Rating
In the October 25 post carried by SiliconANGLE, Musk said:
As of the past few weeks, debt markets have almost entirely shut off access to money for any company that doesn’t have substantial revenue independent of AI data centers. SpaceX closed $40B of debt for GPUs, but we’re investment grade and have Starlink & launch…
CNBC attributed the same points to Musk’s posts, not as its own findings: he said SpaceX was investment grade and generated revenue from Starlink satellite internet and rocket launches, and that debt markets have largely shut off access to financing for companies without substantial revenue outside AI data centers.
The hardware order sits inside a much larger AI unit. CNBC said SpaceX has expanded aggressively into artificial intelligence this year, referring to the business as Super Intelligence, or SI. The segment is now the company’s largest business unit, with third-quarter revenue surging ninefold to $7.56 billion from $813 million a year earlier, according to its latest earnings release.
The Neocloud Shakeout Musk Predicted
SiliconANGLE quoted Musk on the smaller GPU clouds:
200+ neoclouds are going to see guppy-level death and consolidation. Some will make it through, but not many.
Neoclouds rent specialized GPU computing capacity to AI developers and other customers. Many, SiliconANGLE noted, expanded quickly and often used borrowed money to buy the hardware. CNBC reported the same prediction from Musk: more than 200 neoclouds would face widespread failures and consolidation, with only a small number surviving. That is a forecast, not a report that failures are already underway.
- Musk says recent debt access requires revenue independent of AI data centers.
- Neoclouds that scaled on borrowed GPU purchases are the group he flagged.
- Tougher lending could force mergers or canceled expansion, SiliconANGLE said.
- A failure wave could reduce AI chip demand and potentially impact suppliers’ revenue, including Nvidia.
Musk said SpaceX could close the debt because it is investment grade and has Starlink and launch revenue. Companies without substantial revenue independent of AI data centers, he said, have been almost entirely shut off from money.
Hyperscaler Borrowing Beside the SpaceX Package
In June, Morgan Stanley forecast capital expenditure by Amazon Web Services, Microsoft, Google and Meta Platforms at $700 billion this year and over $1 trillion in 2027. Goldman Sachs projected borrowing by those companies at $250 billion this year and $400 billion in 2027. Morgan Stanley said global AI-related debt issuance is expected to top $570 billion this year.
| Capital measure | Amount | Period |
|---|---|---|
| SpaceX GPU debt financing | $40 billion | Chips in Q1 and Q2 2027 |
| Capex: AWS, Microsoft, Google, Meta | $700 billion | This year |
| Capex: those four companies | Over $1 trillion | 2027 |
| Borrowing by those companies | $250 billion | This year |
| Borrowing by those companies | $400 billion | 2027 |
| Global AI-related debt issuance | Over $570 billion | This year |
According to Reuters, as cited by SiliconANGLE, investors are demanding higher yields on AI infrastructure-related corporate bonds because of the volume of new debt. SiliconANGLE said many investors are nervous about whether revenue from AI services will grow fast enough to justify spending on chips, data centers and related infrastructure. Established borrowers are still issuing, at a higher price. Companies without revenue independent of AI data centers, Musk says, have had access to money almost entirely shut off.
