Nvidia has announced a plan to build AI data centers with a whopping $500 billion in funding from major financial companies, including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR. The plan is not just about building new data centers, but also about creating a secondary market for aging GPUs.
To convince these financial companies to commit to the plan, Nvidia has agreed to guarantee the value of its chips used as collateral in these deals. This means that if the chips don’t retain their value as expected, Nvidia will cover up to 25% of the difference. This creates a so-called ‘wrong way’ risk for Nvidia, where its obligations will grow as demand weakens.
The Risks and Rewards
The plan is seen as both brilliant and risky by many. Some have compared it to the failed strategy of Lucent Technologies, which lent its customers money to buy its products during the dotcom bubble. However, Nvidia’s CEO Jensen Huang argues that this plan is different, as the company is bringing in independent, long-term institutional capital to shoulder the bulk of the risk.
Huang has been working to address concerns about circular financing, where Nvidia would be lending money to its customers to buy its products. Instead, the company is getting others to invest in the AI infrastructure market, with Nvidia providing a guarantee on the value of its chips.
A New Ecosystem for AI Hardware
The plan is not just about building new data centers, but also about creating a new ecosystem for AI hardware. Huang envisions a future where AI servers, or ‘AI factories,’ can be used by multiple customers, just like railroads or airlines. This would create a deep market of potential users and offtakers, helping to protect the residual value of Nvidia’s hardware.
This approach could also lead to a broader variety of hardware being used for different AI needs, with startups, enterprises, and researchers tapping into a range of affordable and specialized options. As the leader in the AI market, Nvidia has the power and the opportunity to make this vision a reality.
A Long-Term Strategy
Huang’s plan is a long-term strategy to sustain demand for Nvidia’s hardware, even as the company continues to innovate and release new products. By creating a secondary market for aging GPUs, Nvidia can ensure that its hardware remains valuable and relevant, even as new technologies emerge.
The success of this plan will depend on whether Nvidia can convince investors and customers that its AI hardware is a long-term, investable infrastructure. If the company can achieve this, it could create a new and sustainable market for AI hardware, with Nvidia at the forefront.
Helene Elliott is the senior reporter for News Raise. She covers Science news. She also has a keen interest in photojournalism. Helene holds a nomination for the prestigious Red Smith Award. She is married to author Dennis D’Agostino, a former publicist with the New York Mets.
