Nvidia has teamed up with six of Wall Street’s largest financial firms to create financing platforms aimed at mobilizing more than $500 billion in third-party capital for artificial intelligence infrastructure, the chipmaker announced.The company signed memorandums of understanding with Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs and KKR.
The arrangements are designed to help Nvidia’s customers, including frontier AI labs, enterprises and cloud providers, finance data centers, power generation and the high-end chips needed to train and run advanced AI models.Nvidia CEO Jensen Huang said the partnerships treat computing power as a new investable asset class for the first time. “In AI, compute is revenue,” Huang said.
“We are bringing the world’s leading long-term capital providers together to independently underwrite AI infrastructure.”Under the deals, which remain subject to final agreements, the firms will establish dedicated pools of capital available at attractive rates.
Nvidia has the option to backstop up to 25 percent of the potential financing, or about $125 billion.The move comes as demand for AI computing capacity continues to surge. Big technology companies are projected to spend more than $730 billion this year on AI-related capital expenditures, and the industry faces shortages of both advanced chips and the power and facilities required to run them.
Building the massive “AI factories” that Huang frequently describes requires not only the latest graphics processing units but also reliable electricity, advanced cooling systems and large tracts of land, often in regions where grid capacity is already strained.By bringing private equity firms, asset managers and investment banks into the picture, Nvidia is attempting to turn GPUs and the surrounding infrastructure into assets that institutional investors can underwrite in a manner similar to commercial real estate, toll roads or energy projects.
This approach allows customers to secure large-scale financing without heavily tapping their own balance sheets, potentially accelerating construction timelines that have been slowed by capital constraints and long equipment lead times.Executives from the participating firms described the effort as a practical way to scale digital infrastructure more efficiently. “As we’ve scaled our approach to digital infrastructure, we’ve learned that delivery, not ambition, is the hard part,” KKR co-chief executives Joe Bae and Scott Nuttall said in a joint statement.Nvidia’s GPUs currently power the majority of the world’s leading AI training and inference systems.
The company’s dominance has made it a central player not only in hardware sales but also in shaping how the broader ecosystem finances the next phase of the AI boom. The new platforms are intended to support projects across Nvidia’s customer base, from hyperscale cloud providers to specialized AI startups and enterprise adopters.Industry analysts note that the sheer scale of capital required has outstripped traditional corporate financing methods.
Data center construction costs have risen sharply, power availability has become a limiting factor in several markets, and successive generations of chips continue to demand denser, more expensive facilities. Treating compute as a long-term productive asset is an attempt to attract patient capital that can absorb the multi-year development cycles involved.The announcement underscores Nvidia’s expanding role beyond selling chips.
It is increasingly acting as a central coordinator of the capital, hardware and physical infrastructure required for the AI boom. Shares of Nvidia reacted modestly to the news as investors weighed the scale of the proposed financing against ongoing questions about the pace of AI spending and the timeline for returns on these enormous investments.While the memorandums of understanding mark a significant step, the actual deployment of capital will depend on final deal structures, project pipelines and market conditions. If successful, the initiative could help unlock a new wave of AI infrastructure build-out at a time when the industry is racing to meet rapidly growing demand for computing power.
