Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Erik Bernhardsson highlights how emerging cloud providers like Nscale rapidly build multi-billion-dollar footprints without owning initial hardware. They secure multi-year compute commitments from massive AI labs first, take those signed contracts to Wall Street lenders to finance hardware purchases, and deploy collateralized chips directly to buyers.

Related Insights

Compute providers lock customers into long-term contracts not just for predictable revenue, but because these agreements are prerequisites for securing financing. A bank won't underwrite the construction of a new data center without a multi-year offtake agreement from a credible counterparty, making these contracts foundational to supply growth.

The core innovation of the neocloud model was using firms like Blackstone to finance massive GPU acquisitions. This financial engineering contrasts sharply with hardware startups that spend years on deep R&D. The finance-first approach lowered barriers to entry, leading to a more crowded and competitive landscape.

To hedge against hyperscalers manufacturing in-house silicon, Nvidia actively finances smaller 'NeoClouds' like Firmus. Nvidia acts as a credit guarantor, enabling emerging cloud providers with weak balance sheets to borrow cheaply and purchase tens of thousands of GPUs. In return, Nvidia captures both hardware revenue and ongoing cloud compute revenue cuts.

To finance AI infrastructure without massive equity dilution, firms use debt collateralized by guaranteed, long-term purchase contracts from investment-grade customers. The rapidly depreciating GPUs are only secondary collateral, making the financing far less risky than it appears and debunking common criticisms about its speculative nature.

NVIDIA's new business model involves guaranteeing it will rent back unused GPU capacity from smaller cloud providers. This acts as anchor demand, enabling these 'NeoClouds' to secure financing for massive GPU purchases. It's a strategic move for NVIDIA to build and control its own demand ecosystem, ensuring its chips continue to sell.

Heavy AI infrastructure buyers are adopting a landlord business model to service the massive debt incurred from chip purchases. After borrowing $40 billion to acquire NVIDIA GPUs, SpaceX plans to rent computing capabilities out to rival AI developers like Anthropic and Google for an estimated billion dollars a month. The recurring rental income is specifically structured to cover the interest payments owed on its multi-billion-dollar debt burden.

Silver Lake cofounder Glenn Hutchins contrasts today's AI build-out with the speculative telecom boom. Unlike fiber optic networks built on hope, today's massive data centers are financed against long-term, pre-sold contracts with creditworthy counterparties like Microsoft. This "built-to-suit" model provides a stable commercial foundation.

From cloud providers buying GPUs to companies building data centers, the massive AI buildout is largely financed through debt. This reality means access to compute increasingly depends on a customer's ability to make large upfront down payments and sign long-term contracts, as providers need to secure their own financing.

As the AI build-out matures, financing is shifting from construction to the chips themselves, which can exceed 50% of a data center's cost. Creative solutions are emerging, such as financing backed by the value of the chips or the compute contracts they service, moving beyond traditional loans.

Beyond selling GPUs, Nvidia is providing billions in financial guarantees to smaller "neocloud" companies. This strategic move de-risks data center development for these emerging players, ensuring they can secure debt and build the very infrastructure that will consume Nvidia's chips in the future. Nvidia is effectively underwriting its own future demand.

Neoclouds scale through debt backed by committed customer contracts | RiffOn