SpaceX closes $40 billion debt financing to purchase AI chips

According to Elon Musk, SpaceX has closed $40 billion in debt financing to purchase graphics processing units (GPUs). This once again highlights the growing costs required to expand artificial intelligence computing capacity.

In a post on social media platform X on Sunday, Musk linked the success of the fundraising to the company's investment-grade rating and the revenue generated from Starlink's satellite internet and space launch businesses. However, his statement left unclear exactly what he meant by the deal being "closed," and whether the terms of the debt pricing had been agreed upon with investors.

"This is mostly debt with a buyback right for GPUs, which are expected to be delivered in Q1 and Q2. It won't be a significant factor this quarter," Musk wrote.

According to SpaceX CFO Bret Johnsen, the company expects to spend $35–40 billion on capital expenditures in the fourth quarter.

Earlier this month, Reuters reported that SpaceX was in talks with banks and asset management firms to raise $40 billion, which was to be used to purchase Nvidia's AI chips.

The scale of the financing reflects the enormous capital requirements driven by the rapid growth of the AI industry. Technology companies are competing with one another to acquire advanced processors and build the computing infrastructure necessary for running artificial intelligence systems.

According to Musk, access to debt-market financing has become almost entirely restricted to companies that lack significant revenue independent of the AI data center business. He predicted that consolidation and business failures will become more frequent in the so-called neocloud sector. Later, he also stated that SpaceX has no plans to acquire companies of this type, calling such a move "pointless."

Neocloud companies, including Nebius and CoreWeave, differ from traditional large cloud service providers in that they primarily serve technology clients and offer computing capacity tailored for AI.

Recently, investors have been increasing their bets against US neocloud and data center companies. The reason is growing concern that the rapid expansion of AI infrastructure, largely financed through debt, may prove unsustainable in the long run.