NVIDIA Enlists Six Wall Street Giants to Mobilize $500 Billion for AI Infrastructure
NVIDIA announced on August 10, 2026 that it is partnering with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to create independent financing platforms aimed at mobilizing more than $500 billion in third-party capital for AI compute infrastructure. The partnerships are designed to channel institutional money — from pension funds, sovereign wealth funds, and other large allocators — into the data centers and GPU systems that power AI model training and inference. Inference is the stage where an AI model actually answers queries, as opposed to training, where it learns. NVIDIA's technology stack would serve as the anchor tenant, meaning the foundational component these facilities are built around (NVIDIA News).
What stands out is what NVIDIA is not doing: putting up its own money at that scale. The six financial institutions will act as intermediaries, creating investment vehicles that let outside capital fund AI infrastructure while NVIDIA supplies the architecture, the GPU hardware, and the software ecosystem. The partners bring the distribution network and the ability to raise capital from deep-pocketed investors.
This is the latest in a series of capital-intensive partnerships NVIDIA has assembled in 2026. On July 24, the company announced a $500-billion-plus strategic partnership with SK Group covering AI factories and next-generation memory, formalized at NVIDIA's AI Summit in San Francisco (NVIDIA Investor Relations). In May, NVIDIA and IREN announced a partnership to deploy up to 5 gigawatts of AI infrastructure, with IREN issuing NVIDIA a five-year right to purchase up to 30 million shares at a set price (NVIDIA Investor Relations). In March, NVIDIA committed $2 billion to Nebius as part of a partnership to scale a full-stack AI cloud (NVIDIA News).
The August 10 announcement differs from those earlier deals in one critical respect: the capital is not NVIDIA's. The SK Group, IREN, and Nebius arrangements all involved NVIDIA's own equity or direct investment. The six-firm Wall Street partnership, by contrast, is explicitly designed to mobilize third-party capital. The distinction matters for NVIDIA shareholders because it shifts the financing burden for AI infrastructure off NVIDIA's balance sheet — the company's statement of assets and liabilities — while preserving NVIDIA's position as the indispensable technology provider.
Morgan Stanley said on July 11 that NVIDIA's expanding customer base beyond traditional AI hyperscalers — the handful of giant cloud providers like Amazon, Microsoft, and Google that dominate AI computing — is the key theme for investors to watch (Investing.com). The financing partnerships fit that thesis: Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR collectively manage trillions in assets, and their participation opens a channel into enterprise and institutional AI demand that sits outside the hyperscaler spending cycle. NVIDIA has stated that with its AI factories, financial institutions can develop proprietary banking AI to automate fraud detection, optimize risk, and scale productivity (NVIDIA), meaning several of the six partners are simultaneously capital providers and prospective end customers.
Morgan Stanley itself reported second-quarter 2026 results on July 15 that beat analyst estimates on strong trading and dealmaking, with equity net revenues jumping 69% year over year. The firm is betting on AI investment as a growth driver (Reuters). Goldman Sachs, another of the six partners, has not yet disclosed specific terms of its arrangement with NVIDIA.
The broader context here is a financing gap that has widened as AI infrastructure costs have scaled. A single large AI factory can run into tens of billions in GPU, networking, data center, and power infrastructure costs. Hyperscalers have absorbed the first wave of this spending, but the next phase — enterprise, sovereign, and specialized AI workloads — requires capital sources that sit beyond the balance sheets of the three or four largest cloud providers. The $500 billion target, if realized, would rank among the largest private capital mobilizations for a single technology category.
What remains unaddressed in the announcement is deal velocity and capital deployment timeline. The press materials describe the partnerships as establishing financing platforms; they do not specify committed capital, fund vehicles, or deployment schedules. The $500 billion figure is a mobilization target, not a commitment. For institutional allocators evaluating exposure to AI compute, the key variables — risk-adjusted returns, GPU utilization assumptions, technology obsolescence risk, and competitive dynamics among financing platforms — are not yet defined in public disclosures.
NVIDIA's strategy across these partnerships reveals a deliberate layering: SK Group secures the memory and manufacturing supply chain, IREN addresses power and site development, Nebius extends the AI cloud footprint, and now six Wall Street firms address the capital layer. Each partnership targets a different bottleneck in the AI infrastructure stack, and each keeps NVIDIA's technology at the center without requiring NVIDIA to fund the full buildout.
The question for investors is whether this capital-mobilization model can scale fast enough to meet the demand signals NVIDIA is generating, or whether the gap between announced partnerships and deployed gigawatts widens. The IREN deal's five-year warrant structure gives NVIDIA equity optionality tied to infrastructure delivery, a mechanism that aligns incentives but also signals that buildout timelines extend over years, not quarters. The same will likely hold for the financing platforms announced this week.
