Nvidia has enlisted six of Wall Street's largest alternative asset managers to assemble a $500 billion financing vehicle for artificial intelligence infrastructure, a move that shifts the capital burden for the chipmaker's most expensive customers off corporate balance sheets and onto private credit and insurance money. The lineup includes Apollo Global Management, Blackstone, BlackRock's Global Infrastructure Partners unit, Brookfield Asset Management, Goldman Sachs and KKR, according to a person familiar with the matter who spoke to CNBC on Monday. An announcement could arrive as soon as Monday. The Financial Times first reported the arrangement.
The lineup
The six firms represent a cross-section of the alternative capital ecosystem: three traditional private equity giants in Apollo, Blackstone and KKR, a sovereign-scale infrastructure platform in Brookfield, BlackRock's dedicated infrastructure arm, and Goldman Sachs' balance sheet and advisory franchise. None of the seven companies named responded to requests for comment on Monday. The source spoke on condition of anonymity because they were not authorized to discuss the matter publicly.
The structure
The vehicle is designed to finance the three capital-intensive legs of AI deployment: procurement of Nvidia's high-end GPUs, construction of power-hungry data centers, and long-term electricity capacity contracts. By pooling institutional and insurance capital, the structure allows Nvidia's largest customers, hyperscalers, sovereign wealth funds, and emerging GPU cloud providers, to fund capex without drawing on their own credit capacity or diluting equity. The source did not disclose the split between debt and equity, the targeted leverage, or any break fees or conditions attached to the commitments.
The precedent
Apollo and Blackstone have already structured debt and equity packages for Anthropic, signaling appetite for direct exposure to AI model developers. This effort extends that logic one layer down the stack: financing the physical substrate that makes the models run. Alternative managers have been chasing digital infrastructure yield for years; the AI boom has widened the addressable market from fiber and towers to gigawatt-scale compute campuses.
What to watch
Whether the $500 billion figure represents committed capital, a fundraising target, or a notional capacity ceiling remains unclear. The speed of deployment, and the terms offered to borrowers, will reveal whether this is a genuine new capacity source or a marketing coordination exercise. If the vehicle prices risk at spreads that reflect the opacity of AI demand forecasts, it becomes a subsidy to Nvidia's revenue growth dressed up as private market discipline.