Urgent.News

What's breaking now, across thousands of outlets.

AI

Elon Musk issues a warning for 'AI neoclouds' as SpaceX borrows $40 billion for chips

SpaceX has raised $40 billion in debt to buy GPUs, while Musk predicts a major shakeout among smaller AI cloud providers.

Elon Musk issues a warning for 'AI neoclouds' as SpaceX borrows $40 billion for chips

SpaceX, the private aerospace company led by Elon Musk, has borrowed $40 billion in debt to purchase GPUs, powerful processors utilized in artificial-intelligence systems. The company announced this deal on X, alongside a warning for neoclouds—companies whose revenue relies primarily on data centers—potentially exacerbating AI concerns on Wall Street and reducing risk tolerance in broader financial markets, including cryptocurrencies.

The $40 billion debt is primarily callable, meaning SpaceX can repay it early. These funds could be used to cover GPU deliveries anticipated in the first half of 2027, though Musk did not explicitly confirm this. Prior to this, Reuters reported that SpaceX was in discussions with banks and asset managers to raise $40 billion for purchasing Nvidia AI chips.

Elon Musk explained that SpaceX's creditworthiness and consistent revenue from Starlink internet service and rocket launches allowed the company to secure this financing. In contrast, firms that rely entirely on AI data centers for revenue are facing difficulty in accessing such funds, as debt markets have largely closed off to companies lacking independent revenue sources.

Musk pointed out that as of recent weeks, debt markets have almost entirely withdrawn access to money for entities without substantial revenue independent of AI data centers.

Musk predicted that over 200 "neoclouds"—new companies renting out AI computing power—will consolidate, leaving only a few survivors. This phenomenon underscores the escalating costs associated with the AI computing race, as constructing large clusters of AI chips often requires tens of billions of dollars. While companies with diverse revenue streams and strong credit ratings can still borrow, those whose business fundamentally depends on AI infrastructure face more stringent lending conditions due to lenders' increasing selectivity.

Venture capitalist Chamath Palihapitiya had previously cautioned that many neoclouds appear underfunded and poorly managed. He argued that technology giants Amazon, Microsoft, and Google are moving cautiously in these deals, preferring to retain computing power for their own use. He suggested that delays in building data centers could result in many chips and incomplete projects either sitting idle or failing to generate sufficient revenue to cover financing costs by 2027.

Recently, financial markets have shown heightened apprehension regarding the returns on substantial AI investments, which has adversely affected AI-linked stocks and the overall market. As of this report, Nasdaq futures were trading near 29,080, a significant decline from above 30,000 earlier in the month. Additionally, diversified real-world asset stablecoins offer yields ranging from 5-7% through real credit, while crypto funding has decreased to approximately 4%.

Similarly, GENIUS stablecoins expand their yield off-chain, with a target market capitalization (TAM) of $4 billion within three years.

Written by urgent.news from CoinDesk's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at coindesk.com →

More in AI

How to Stop Claude From Sounding Like AI (My Claude Project Instructions)

Every time I opened a new chat in Claude, I used to waste the first ten minutes fixing the exact same problem. I would explain what I do, tell it who I was writing for, and ask for a normal…

  • Claude's initial output sounds stiff due to lack of specific context
  • Create 00MYCONTEXTMEMORY.md with work background, audience, offerings, results, and business goal
  • Add 01HUMANVOICERULES.md with plain-English writing guidelines to Project

More from Monday 26 October →