The "Gold Rush Era" of American Data Centers

The AI wave is pushing the U.S. data center industry into a capital frenzy. Funds and new players are flooding in, and innovative financing structures are emerging one after another. However, beneath the surface of prosperity, the huge gap between profit expectations and reality, the fragility of cyclical dependencies, and the inexperience of new entrants are accumulating systemic risks in this gold rush.

On October 20th, according to technology media The Information, the atmosphere at last week's data center industry conference in Las Vegas was drastically different from a year ago. OpenAI, xAI, and Meta have pledged to invest hundreds of billions of dollars over the next decade, shifting the focus of discussions from "difficulty in finding land and electricity" to "who can build the most gigawatt capacity" data centers. An investment group led by BlackRock and MGX acquired the 12-year-old data center operator Aligned Data Centers for a record $40 billion.

Reports indicate that behind the optimistic sentiment lie real challenges. Taking Oracle as an example, its actual financial data for the past five quarters in the AI cloud business shows that the current profit margin from leasing NVIDIA chips is 15-20 percentage points below the target. Industry insiders express caution in private, warning of "excessive circular funding" or "over-reliance on a single company" in transaction structures.

Innovative Financing Structures Become the New Normal

To support astronomical investments, the industry is inventing various creative financing methods.

Leaseback transactions have become a new favorite, with xAI purchasing NVIDIA chips from its main investor Valor Equity Partners and then leasing them back for use. OpenAI is also discussing a similar structure with NVIDIA—developing and managing data centers itself but reducing costs through leaseback to avoid paying premiums to Oracle and Microsoft.

Reports suggest that the essence of these transactions is a risk-sharing mechanism that blurs the boundaries between customers, suppliers, and financiers, allowing funds to continuously flow into data center construction. The acquisition of Aligned Data Centers acts like a shot of adrenaline, motivating more operators to seek buyers.

At the same time, NVIDIA is not only a chip supplier but is also deeply involved in the financing aspect—providing financing for chip customers and data center projects, with funds ultimately flowing back in the form of chip purchases.

Industry insiders are concerned whether this circular flow of funds distorts real demand and whether NVIDIA's dual role as both referee and player could create a market bubble. However, OpenAI's recent commitment to using AMD chips and chips co-designed with Broadcom shows an intention to break NVIDIA's monopoly.

AI Companies Cross-Industry Challenge Industry Rules

In this "gold rush" of data centers, the most striking phenomenon is the role displacement.

Reports indicate that Poolside, originally an AI programming startup, now claims to be building a 2-gigawatt data center, planning to lease part of it to AI cloud service provider CoreWeave, and asserts that it has cracked the industry's most pressing bottleneck. Startups like Fermi are directly jumping into multi-gigawatt projects, betting they can outperform cloud computing giants like Google and Microsoft in speed and performance These newcomers, lacking traditional data center development experience, are challenging the existing rules of the industry. Traditional data center developers are increasingly skeptical of the capabilities of new players. Microsoft executives have expressed to OpenAI that they do not believe Oracle can deliver on its promised thousands of megawatts of capacity.

Several industry insiders have been recruited by newcomers to "solve urgent operational challenges." Many predict an impending reshuffling moment, where overly ambitious projects will collapse due to delays, power shortages, or unrealistic timelines.

However, profit realities test business models

Behind the optimistic sentiment lies a reality of challenges. Oracle provided optimistic revenue and profit margin forecasts at its annual cloud conference, but the actual financial data from the past five quarters reveals a harsh truth—current profit margins for leasing NVIDIA chips are 15-20 percentage points below target values.

AI cloud service providers are in a race against time: they must purchase expensive NVIDIA chips in advance, but customers only start paying once projects are completed and performance standards are met. Uncontrollable factors such as power supply and equipment delays can disrupt plans at any moment.

When the identities of suppliers, customers, and financiers overlap, systemic risks accumulate.

Industry leaders are privately wary of this cycle of dependency's fragility. When Microsoft chose to let Oracle take on part of OpenAI's server needs, the industry's shrewdest players were already speaking through their actions: either they are pessimistic about long-term demand or unwilling to take on excessive risks.

Analysis indicates that in this gold rush, NVIDIA firmly holds the "shovel-selling" position, traditional cloud giants have technological accumulation and risk tolerance, while newcomers face the greatest uncertainty. Only those with real technical capabilities, ample financial reserves, and risk management experience can stand firm when the tide recedes