Chip fever has created an $11 billion debt market backed by Nvidia’s AI-enabling GPUs

• Top Wall Street lenders have created a big debt market back by red-hot Nvidia GPUs.
• A report from the FT says firms have made billions in loans to companies that use Nvidia chips as collateral.
• The risk is that new GPU technology is released regularly, and the existing collateral could lose value.

Top Wall Street firms have loaned billions to a handful of tech companies that have used Nvidia’s artificial intelligence-enabling GPUs as collateral, The Finanacial Times reported.

According to the FT, big financial institutions including BlackRock, Blackstone, Pimco, and Carlyle, have made over $11 billion in loans to “neocloud” companies, tech companies that provide cloud services to other tech firms building AI products.

These firms are among some of Nvidia’s largest customers and include names like CoreWeave, Crusoe, and Lambda Labs. Collectively, they have purchased tens of thousands of Nvidia’s highly sought-after chips and have offered them as collateral for loans that they use to buy even more GPUs, the report said.

There are concerns, though, over the value of Nvidia’s GPUs, with the price of existing AI chips potentially coming down as more advanced models are released and as companies potentially dial back AI spending.

Having access to GPUs is no longer like having a “golden ticket to Willy Wonka’s factory,” as it was considered to be a year ago, a senior executive at one of CoreWeave’s lenders told the FT.

Leasing contracts between neocloud firms and tech groups, meanwhile, are poised to expire in the coming years. That could result in a huge supply of chips making their way back to the market, the report added.

CoreWeave, Crusoe, and Lambda Labs did not immediately respond to a request for comment from Business Insider. Nvidia declined to comment.

Elsewhere on Wall Street, there are concerns that Nvidia’s huge pace of growth cannot be sustained, with doubts rising about returns on AI investment by corporations and the threat of competition from other chip makers muscling into the AI space.