NVIDIA announced partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to create independent financing platforms aimed at mobilizing more than US$ 500 billion in third-party capital to build artificial intelligence infrastructure.
The amount represents aggregate third-party capital that the platforms can mobilize over time, not NVIDIA revenue, a single fund, or a commitment to a single client. The financial institutions will independently evaluate each opportunity, including client, demand, utilization, cash flow, and residual value.
Financing model
The manufacturer will provide the AI factory platform, while the partners will contribute long-term capital and financing expertise. In some cases, NVIDIA may offer a residual value support mechanism of up to 25% of the value of an opportunity, assessed on a project-by-project basis.
NVIDIA said the goal is to expand access to capital for AI labs, companies, and cloud providers that have demand for computing but lack financing at the required scale. Capacity will be built based on real customer economics.
Demand and prices
NVIDIA highlighted that AI computing has come to be treated as a productive asset. The A100 chip, launched in 2020 with the Ampere architecture, remains in active commercial use for training, fine-tuning, inference, and high-performance computing, with multi-year contracts.
Rental prices have risen: the one-year rental of H100 went from about US$ 1.70 per GPU hour in October 2025 to US$ 2.35 in March 2026. The median on-demand price among providers rose from US$ 2.00 per hour in October 2025 to US$ 2.70 in June 2026. Blackwell capacity is sold at a premium, with cloud rates for B200 between US$ 5.30 and US$ 7.05 per hour.
