Nvidia seeks Wall Street partnerships to finance AI boom with $500B capital target

Nvidia, the world’s most valuable public company at roughly $5.3 trillion, has decided that selling the most sought-after chips on the planet isn’t enough. Now it wants to become the matchmaker between Wall Street’s mountains of capital and the AI industry’s insatiable appetite for data centers.

The company announced memorandums of understanding with six heavyweight financial firms: Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs, and KKR. The goal is to create independent “compute financing platforms” designed to funnel more than $500 billion in third-party capital toward AI infrastructure like data centers and chip factories.

How the financing structure works

Nvidia wants to help treat AI compute hardware and infrastructure the way investors already treat toll roads, pipelines, and power plants: as long-duration, revenue-generating assets that can be financed through institutional capital.

Under this framework, Wall Street firms would finance the hardware and facilities, then lease compute capacity to frontier AI labs, enterprises, and cloud providers. The financial firms earn steady returns. The AI companies get access to compute without enormous upfront capital outlays. Nvidia sells more chips.

Nvidia itself is not putting cash on the table or taking on debt. Instead, the company is offering optional residual-value support of up to 25% on select projects. That’s a form of backstop where Nvidia essentially guarantees that the hardware retains a portion of its value over time, giving lenders more confidence in the collateral.

Each of the six partners would establish its own independent platform, meaning this isn’t a single mega-fund but rather a constellation of competing financing vehicles all oriented around the same asset class.

Why Nvidia is playing financial architect

The $500 billion target reflects the scale of investment that industry leaders believe is necessary. For context, that figure exceeds the annual GDP of countries like Norway or Austria.

From Nvidia’s perspective, the logic is circular in the best possible way. More available financing means more data centers get built. More data centers mean more GPU orders. More GPU orders mean more Nvidia revenue. The company doesn’t need to lend a dollar to benefit enormously from this ecosystem.

The risks nobody wants to talk about

As of late September 2026, no capital has actually been raised under these new platforms. The partnerships remain in the memorandum-of-understanding phase, meaning final agreements haven’t been signed.

Nvidia’s residual-value guarantees add another layer of complexity. Even though the support is optional and capped at 25%, it creates an indirect financial obligation. If AI hardware depreciates faster than expected, perhaps because of a new chip architecture that renders current GPUs less competitive, Nvidia could face claims on that backstop. Market analysts are watching this dimension closely.

Blackstone and Brookfield have already been building significant positions in data center infrastructure through direct acquisitions. Goldman Sachs has been expanding its infrastructure lending. These firms aren’t entering unfamiliar territory so much as formalizing a trend that was already underway.