CoreWeave, an NVIDIA-backed cloud service provider specializing in GPU-accelerated services, has secured a debt facility worth $2.3 billion using NVIDIA's H100-based hardware as collateral. The company intends to use the funds to procure more compute GPUs and systems from NVIDIA, construct new data centers, and hire additional personnel to meet the growing needs for AI and HPC workloads.

CoreWeave has reaped enormous benefits from the rise in generative AI due to its large-scale cloud infrastructure as well as an exclusive relationship with NVIDIA, and its ability to procure the company's H100 compute GPUs as well as HGX H100 supercomputing platforms amid shortages of AI and HPC hardware. Since many AI and HPC applications used nowadays were developed for NVIDIA's CUDA platform and API, they require NVIDIA's GPUs. Therefore, access to H100 gives CoreWeave a competitive edge over traditional CSPs like AWS, Google, and Microsoft.

In addition to offering its customers access to advanced hardware, CoreWeave collaborates with AI startups and major CSPs — which are essentially its competitors — to build clusters that power AI workloads. These rivals — AWS and Google — have their own processors for AI workloads, and they continue to develop new ones. Still, given the dominance of CUDA, they have to offer NVIDIA-powered instances to their clients and are currently grappling with NVIDIA GPU supply limitations.

CoreWeave's competitive advantage, facilitated by access to NVIDIA's latest hardware, is a key factor in the company's ability to secure such substantial credit lines from companies like Magnetar Capital, Blackstone, Coatue, DigitalBridge, BlackRock, PIMCO, and Carlyle. Meanwhile, CoreWeave has already gotten $421 million from Magnetar at a valuation exceeding $2 billion.

Notably, this is not the first example of an NVIDIA-supported startup reaping substantial benefits from its association with the tech giant. Last month, Inflection AI built a supercomputer worth hundreds of millions of dollars powered by 22,000 NVIDIA H100 compute GPUs.

Meanwhile, this is the first time NVIDIA's H100-based hardware was used as collateral, emphasizing these processors' importance in the capital-intensive AI and HPC cloud business. Moreover, this massive loan indicates the growing market for private asset-based financing secured by actual physical assets.

"We negotiated with them to find a schedule for how much collateral to go into it, what the depreciation schedule was going to be versus the payoff schedule," said Michael Intrator, CoreWeave's CEO. "For us to go out and to borrow money against the asset base is a very cost-effective way to access the debt markets."

The company recently announced a $1.6 billion data center in Texas and plans to expand its presence to 14 locations within the U.S. by the end of 2023.

Source: Reuters

Comments Locked


View All Comments

  • Threska - Friday, August 4, 2023 - link

    Power of a monopoly. Great for them, not so great for those doing AI and HPC. Might explain why the biggest are making their own AI chips.
  • PeachNCream - Friday, August 4, 2023 - link

    Hope it works out for them, but going into billions of debt with components that are almost obsolete as they leave the fab is certainly playing with fire.
  • name99 - Friday, August 4, 2023 - link

    That's kinda the point. Under normal circumstances, these would not be great collateral; but circumstances are not normal and enough people believe that these cards are valuable today and will remain valuable for a few years, to be willing to provide the loan.

    In other words this is the COMPLETE OPPOSITE of "almost obsolete as they leave the fab"...
  • PeachNCream - Sunday, August 6, 2023 - link

    People that believe in the value of a product on an 18 month upgrade cycle willing to do a couple billion into debt on hardware - nevermind the place for that hardware to live - are unrealistic.
  • brucethemoose - Friday, August 4, 2023 - link

    > almost obsolete as they leave the fab

    A100s are basically the standard. Most public models and finetunes you see going around were trained on A100s, and many ML papers still cite A100s.

    Researchers, frameworks and stuff seem to be transitioning to the H100 just this year. Heck, I think it needs CUDA 12 for good utilization, and thats not even part of a stable PyTorch release yet.
  • FinancePerson - Wednesday, August 9, 2023 - link

    This is the key point. A100s have been available since 2020 and still 3 years later it's almost impossible to find them available on a public cloud. Their utilization is maxed. Given most depreciation schedules are 3 or 4 years, that means these CoreWeave assets will be in-the-money with plenty of residual life once they're paid off.

    I find it funny the way most journalists are depicting this story. This form of ABL is very common in infrastructure investing. Structured finance is a specialized field that allows debt investors to securitize the cash flows of a multi-year contract, especially when its a hyperscaler (most likely Microsoft). This is nothing new folks. Saying the "GPUs are collateral" is not the whole truth, it's the entire infrastructure and contracts surrounding them.
  • waterdog - Saturday, August 5, 2023 - link

    "Hope it works out for them, but going into billions of debt with components that are almost obsolete as they leave the fab is certainly playing with fire." That's the creditors' problem, not CoreWeave's problem.
  • PeachNCream - Sunday, August 6, 2023 - link

    Typical thinking of the indebted.
  • boozed - Sunday, August 6, 2023 - link

    To paraphrase Getty; if you owe the bank $2.3 billion, that's the _bank's_ problem.
  • brucethemoose - Friday, August 4, 2023 - link

    A rumor suggests Nvidia is cutting back Lovelace gaming GPU production to make more server GPUs.

Log in

Don't have an account? Sign up now