Imagine betting billions of dollars on technology that might be outdated in just three years. That’s exactly what’s happening in Silicon Valley’s AI arms race, where tech companies are turning to creative financing schemes to fund their massive chip purchases. As demand for AI computing power skyrockets, companies are increasingly using loans backed by the very chips they need to train their models – creating a high-stakes financial ecosystem where yesterday’s breakthrough could become tomorrow’s paperweight.
The GPU-Backed Loan Boom
Tech companies are increasingly turning to loans backed by the graphics processing units (GPUs) on which their large language models are trained as they hunt for ways to fund their massive AI investments. These loans, secured against chips and backed by leases to tech groups, have become popular with a sector burning hundreds of billions of dollars a year in the AI arms race on hardware that can quickly become obsolete.
“Investors are very excited,” said David Ridenour, a partner specializing in finance and restructuring at law firm King & Spalding. “People are willing to dive into [GPU deals] on a take-it-or-leave-it basis.” Pioneered by cloud computing provider CoreWeave in late 2023, GPU-backed debt has grown in popularity as demand for advanced chips skyrockets and prices soar. Citigroup estimates that GPUs and associated servers can account for 30 to 40 percent of total project costs for data centers.
The Nvidia Juggernaut and Market Dynamics
This financing frenzy comes as Nvidia, the dominant player in AI chips, reports staggering growth. The company recently announced quarterly revenue of $68.1 billion, up 73% from a year ago, with data center revenue reaching $62.3 billion. CEO Jensen Huang highlighted exponential growth in computing demand, stating, “Computing demand is growing exponentially… Our customers are racing to invest in AI compute.”
But Nvidia’s success tells only part of the story. The company now faces scrutiny over circular financing deals and geopolitical tensions affecting chip sales to China. Meanwhile, competitors are emerging: Meta recently announced a multiyear agreement to purchase up to $100 billion worth of AMD chips as part of its strategy to diversify AI compute infrastructure. This diversification push highlights how tech giants are trying to reduce their reliance on any single chip supplier while still feeding their AI ambitions.
The High-Stakes Financing Structure
The loans work through a clever financial structure: tech companies and investment firms form special-purpose vehicles to acquire caches of high-performing chips, which are then leased back to the tech businesses. This arrangement allows Big Tech groups to shift the loans off their corporate balance sheets while securing the computing power they desperately need.
Recent deals illustrate the scale of this trend. Apollo announced a $3.5 billion financing package for a digital infrastructure fund that would buy Nvidia’s GB200 hardware and lease it to Elon Musk’s xAI. IREN Limited secured $3.6 billion in loan commitment from Goldman Sachs and JPMorgan to buy chips for its AI contracts with Microsoft. Lenders often have to act fast – “A big player would basically ask, ‘would you like to participate in a deal that closes in two weeks and throw in a couple hundred million?'” said a lawyer familiar with GPU financing.
The Obsolescence Dilemma
Here’s where the gamble becomes clear: deals usually come with “hell or high water” clauses that prevent tech companies from terminating leases early, helping mitigate the risk that these GPUs become obsolete as AI technology quickly evolves. But some investors remain deeply concerned.
“It is a very new space and a lot of people are grappling with the question of GPU lifespan,” said Dorina Yessios, US co-head of energy, infrastructure and natural resources at A&O Shearman. “That has to be factored into underwriting, just like any other equipment financing.”
The skepticism runs deeper among some investors. “Those things won’t make it three years before they are antiquated. It’s a huge gamble,” said an investor who has turned down multiple GPU financing pitches. “The idea of reselling GPUs from a few years ago [after a default] is like beating a dead horse.”
The Credit Risk Perspective
This financing boom occurs against a backdrop of broader credit market concerns. UBS credit strategist Matthew Mish has analyzed a potential rapid, severe AI disruption scenario that could trigger a credit crunch. The scenario projects significant default rate increases in high-yield bonds (3-6%), leveraged loans (8-10%), and private credit (14-15%), with spreads widening substantially.
“Investors increasingly want to talk about AI disruption and our tail risk scenario: a rapid, severe AI disruption,” Mish noted. While this isn’t UBS’s baseline scenario, it highlights how financial markets are beginning to price in the potential volatility of the AI investment boom.
The Business Reality Check
Despite these risks, the AI compute race shows no signs of slowing. Nvidia’s Huang argues that compute investments will soon bring revenue: “In this new world of AI, compute is revenue. Without compute, there’s no way to generate tokens. Without tokens, there’s no way to grow revenues.”
But the question remains: Are tech companies and their financiers building sustainable infrastructure or creating a bubble of rapidly depreciating assets? The answer may determine not just which companies win the AI race, but how the entire financial ecosystem supporting this technological revolution weathers the inevitable waves of innovation and obsolescence.
As one investor put it when discussing GPU lifespan concerns: “We really want to ensure the GPUs’ useful life well exceeds the amortised period of our investment.” In an industry where technological advancement happens at breakneck speed, that’s becoming an increasingly difficult promise to keep – and an increasingly expensive bet to make.