Crusoe and other 'neoclouds' lead the charge in the U.S.'s ambitious AI data center initiative, creating jobs and innovation.
 |
| AI-focused startups gain momentum in a $500 billion U.S. initiative to stay ahead in the global race for artificial intelligence. Illustration: ChicHue |
Washington, USA — Jan 24, 2025:
A new $500 billion private-sector initiative to build AI-focused data centers in the United States is putting the spotlight on smaller cloud computing firms that are challenging industry giants. The project, announced this week by former President Donald Trump, is expected to create over 100,000 jobs and help the U.S. maintain its lead in the global AI race, particularly against China.
Key players in the initiative include OpenAI, SoftBank, and Oracle. However, sources familiar with the project revealed that San Francisco-based startup Crusoe has been tapped to construct the first data center under the plan, codenamed Stargate. Oracle reportedly partnered with Crusoe to expedite construction and reduce costs, then began discussions with OpenAI as a potential customer. Oracle has not commented on the reports.
Crusoe is part of a new breed of cloud providers known as "neoclouds," which specialize in meeting the unique demands of AI workloads. Other companies in this group include CoreWeave, Nebius Group, and Lambda. Unlike major cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud, neoclouds focus exclusively on AI clients and rely heavily on Nvidia chips rather than developing proprietary alternatives.
“Smaller companies can move faster,” said Robert Brooks IV, vice president of revenue at Lambda, referring to the Stargate project. “While big players like AWS or Google Cloud often focus on broader, non-AI-related considerations, neoclouds are designed to prioritize AI needs.”
Neocloud firms have also become key partners for Nvidia, providing a growing market for its AI chips. Stacy Rasgon, a semiconductor analyst at Bernstein, noted that this trend strengthens Nvidia's position. “It looks like Nvidia just gained another major customer,” Rasgon said.
Despite their agility, neoclouds face challenges in scaling their operations to meet the demands of large-scale projects like Stargate. Crusoe is building a data center in Abilene, Texas, with plans to house 100,000 specialized AI chips per building. To speed up construction, the company is using pre-fabricated components, a method similar to modular home construction.
Other neoclouds, however, rely on leasing existing facilities or collaborating with data center specialists like Equinix and Compass to fast-track their development. Crusoe has already handed over its first building for tenant improvements in just six months, a process that typically takes years.
Meanwhile, CoreWeave is preparing for an initial public offering, reportedly targeting a valuation of $35 billion, as it positions itself as a leading player in the growing AI cloud market.
The project’s ambitious scale has raised questions about its financing. Elon Musk recently expressed skepticism about whether SoftBank and OpenAI can fully fund the initiative, with reports suggesting that a mix of debt financing and direct contributions—estimated at $19 billion per company—may be required.
The AI data center initiative has the potential to transform the cloud computing landscape. Jason Hardy, Chief Technology Officer for AI at Hitachi Vantara, highlighted the scope of the effort, describing it as unprecedented in scale and a significant driver of job creation. However, he cautioned that its long-term impact remains uncertain.
As startups like Crusoe and CoreWeave race to meet the demands of this initiative, they are proving that speed and specialization can rival the dominance of established tech giants, reshaping the competitive dynamics of the AI-driven cloud market.