Wall Street Capital Moves Into Nvidia AI Infrastructure

More than $500 billion of third-party capital could support AI infrastructure through independent financing platforms established by Nvidia (Nasdaq: NVDA) and six financial institutions, Nvidia announced the partnerships on Aug. 10. The partners are Apollo, Blackrock, Blackstone, Brookfield, Goldman Sachs and KKR.

Jensen Huang, founder and CEO of Nvidia, said:

“NVIDIA has reached an important milestone. We began by building chips; today, we are helping create a new class of productive, investable infrastructure: AI factories.”

“In AI, compute is revenue. NVIDIA compute is uniquely suited for this role,” the executive stated.

The partners represent some of the world’s largest asset managers, private-capital firms and investment banks. Blackrock, the world’s largest asset manager with $15.3 trillion in assets under management, anchors the group, alongside Blackstone with more than $1.3 trillion in assets under management, Apollo with about $1.05 trillion and Brookfield with more than $1 trillion. Goldman Sachs is a global investment bank, and KKR is an alternative asset manager with businesses spanning private equity, credit, infrastructure and capital markets.

The firms signed memorandums of understanding to create dedicated capital pools for Nvidia customers, including frontier AI labs, enterprises and AI cloud providers. The proposed partnerships remain subject to final agreements.

How the $500 Billion Financing Would Work

Financial institutions will independently evaluate customers, demand, utilization, cash flow, and residual equipment value before financing projects. The $500 billion represents aggregate third-party capital the platforms are designed to mobilize over time, not Nvidia revenue or a single committed fund.

Wall Street is already expanding AI lending, with global AI-related debt issuance projected to reach $570 billion in 2026, alongside data-center capital expenditures tracking near $850 billion.

The structure also addresses concerns that chipmakers could effectively finance customers buying their own hardware, potentially inflating reported demand. Huang addressed the concern in an Aug. 10 post on X, describing Nvidia’s potential support as limited, residual-value based and designed to complement independent underwriting rather than replace it.

Nvidia may provide a residual-value support mechanism for up to 25% of an individual financing opportunity, assessed project by project rather than across the full $500 billion initiative. The financial institutions would still determine which projects receive financing based on customer demand, utilization, cash flow and residual value.

Major AI Projects Are Already Drawing Capital

Nvidia plans to invest $1 billion in Naver, while Brookfield has entered a nonbinding term sheet to provide up to $9 billion for expanded AI infrastructure in South Korea.

SharonAI Holdings Inc. (Nasdaq: SHAZ) signed a six-year Nvidia collaboration on June 12 covering 72 megawatts of Australian data-center capacity and as many as 40,000 Grace Blackwell GB300 GPUs. The agreement has a contract value of up to $4.88 billion and uses a revenue-sharing and credit-support model.

Contracted AI revenue is also appearing beyond the largest technology companies. HIVE Digital Technologies Ltd. (Nasdaq: HIVE) signed a three-year AI cloud agreement worth about $220 million involving Bell Canada and Cohere.

Blackrock Chairman and CEO Larry Fink said:

“The AI buildout will require unprecedented investment and a skilled workforce to turn that investment into the infrastructure that will help power future growth.”

“Together, we can help deliver the compute capacity that companies need to grow and create more jobs, supporting the continued growth of the U.S. and global economies, while creating attractive, long-term investment opportunities for our clients,” Fink added.

Nvidia Wants GPUs Treated as Long-Lived Assets

Huang argues Nvidia compute can retain value through redeployment across customers and software improvements from CUDA. Nvidia’s A100, introduced in 2020, remained commercially active six years later. Decentralized physical infrastructure networks are also using GPU compute as an on-demand infrastructure resource, with networks such as Render, Akash and io.net connecting computing capacity with users.

Goldman Sachs has identified AI chip longevity as a key factor in infrastructure costs, with useful silicon life influencing its $7.6 trillion AI capital-spending projection.

Rising rental prices also point to strong demand for Nvidia’s AI chips. One-year H100 rental rates rose from about $1.70 per hour in October 2025 to roughly $2.35 in March, while newer B200 chips rented for about $5.30 to $7.05 per hour.