Nvidia and Wall Street Partners Target $500 Billion for AI Infrastructure
Nvidia and a group of major Wall Street firms said Monday they are aiming to channel at least $500 billion into artificial intelligence infrastructure, a broad push to bring more institutional money into data centers powered by Nvidia systems.
The effort brings together Goldman Sachs and investment firms Apollo, BlackRock and KKR in what Nvidia described as a partnership to raise funds for AI-focused data center projects. Nvidia did not specify the timeline for reaching the $500 billion target.
The plan is designed to tap growing investor appetite for AI exposure as cloud providers including Amazon, Google, Microsoft and Meta continue to pour hundreds of billions of dollars into data center expansion. Nvidia, the dominant supplier of chips used to train and run AI models, has become a central player in that buildout, and the new funds could extend its reach to a wider range of AI research labs and startups.
Nvidia chief executive Jensen Huang told CNBC that the initiative could give AI companies access to the capital they need to scale. Data center construction, he said, requires roughly $50 billion to $60 billion in investment for each gigawatt of power capacity. BlackRock chief executive Larry Fink said current estimates suggest the United States alone will need more than 70 gigawatts of AI data center capacity to keep up with demand.
Goldman Sachs chief executive David Solomon acknowledged the AI race would likely produce winners and losers. Even so, the firms involved expect computing capacity to become a more established investment category over time.
The fundraising plan also highlights a growing debate over how AI-related investments are being structured. Nvidia has invested in several AI companies in recent months, and some of those companies have used the proceeds to buy Nvidia products. Some investors have raised concerns about the circular nature of those deals, adding a note of caution to an otherwise aggressive push.
If the $500 billion target is met, the initiative would represent one of the largest coordinated efforts to treat AI infrastructure as a standalone asset class, with implications for power markets, construction, and the broader technology economy.