is a reporter who writes about tech, money, and human behavior. She joined The Verge in 2014 as science editor. Previously, she was a reporter at Bloomberg.
The AI data center build-out, as it currently stands, is dependent on two things: Nvidia chips and borrowed money. Perhaps it was inevitable that people would begin using Nvidia chips to borrow money. As the craze has gone on, I have begun to worry about the weaknesses of the AI data center boom; looking deeper into the financial part of this world, I have not been reassured.
Nvidia has plowed plenty of money into the AI space, with more than 70 investments in AI companies just this year, according to PitchBook data. Among the billions it’s splashed out, there’s one important category: neoclouds, as exemplified by CoreWeave, the publicly traded, debt-laden company premised on the bet that we will continue building data centers forever. CoreWeave and its ilk have turned around and taken out debt to buy Nvidia chips to put in their data centers, putting up the chips themselves as loan collateral — and in the process effectively turning $1 in Nvidia investment into $5 in Nvidia purchases. This is great for Nvidia. I’m not convinced it’s great for anyone else.
Do you have information about loans in the AI industry? You can reach Liz anonymously at lopatto.46 on Signal using a non-work device.
There has been a lot of talk about the raw technical details of how these chips depreciate, and specifically whether these chips lose value so fast they make these loans absurd. While I am impressed by the sheer amount of nerd energy put into this question, I do feel this somewhat misses the point: the loans mean that Nvidia has an incentive to bail out this industry for as long as it can because the majority of GPU-backed loans are made using Nvidia’s own chips as collateral.
Of course, that also means that if something goes wrong with Nvidia’s business, this whole sector is in trouble. And judging by the increasing competition its chips face, something could go wrong soon.
Can startups outrun chip depreciation — and is it happening faster than they say?
Loans based on depreciating assets are nothing new. For the terminally finance-brained, products like GPUs register as interchangeable widgets (in the sense of “an unnamed article considered for purposes of hypothetical example,” not “gadget” or “software application”) not substantively different from trucks, airplanes, or houses. So a company like CoreWeave can package some chips up with AI customer contracts and a few other assets and assemble a valuable enough bundle to secure debt, typically for buying more chips. If it defaults on the loan, the lender can repossess the collateral, the same way a bank can repossess a house.
One way lenders can hedge their bets against risky assets is by pricing the risk into the interest rate. (There is another way of understanding debt, and we will get there in a minute.) A 10-year mortgage on a house is currently 5.3 percent. CoreWeave’s first GPU-backed loan, made in 2023, had 14 percent interest in the third quarter of this year. (The rate floats.)
“You have so many forces acting in making them a natural monopoly, and this amplifies that.”
... continue reading