NVIDIA Promotes AI Factories as Investable Assets

NVIDIA has detailed its economic argument for classifying its AI factory compute as an investable infrastructure asset class. This move follows the signing of memorandums of understanding on August 10, 2026, with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR. These agreements aim to establish independent financing platforms to mobilize more than $500 billion in third-party capital for AI infrastructure. To support these projects, NVIDIA disclosed it may provide residual-value support for up to 25% of an opportunity on a project-by-project basis.

The chipmaker argues that its DSX AI factories behave like revenue-generating utilities rather than depreciating IT hardware. NVIDIA points to rising GPU rental rates to prove this durability. For instance, one-year H100 rental pricing climbed from $1.70 per GPU-hour in October 2025 to $2.35 in March 2026. Meanwhile, cross-provider on-demand median rates rose from $2.00 to $2.70 over a similar period, and Blackwell B200 cloud rates command a premium between $5.30 and $7.05 per GPU-hour. Additionally, older hardware like the Ampere-based A100, launched in 2020, remains in active commercial use six years later.

The six participating financial institutions will independently evaluate and finance each project. These partners hold massive capital reserves; as of June 30, 2026, Apollo managed approximately $1.05 trillion in assets, Blackstone oversaw more than $1.3 trillion, and Brookfield managed over $1 trillion. NVIDIA asserts this independent underwriting structure addresses concerns about circular demand, as the capital comes from external institutions targeting diverse buyers like frontier AI labs, startups, and governments.

For AI practitioners, this shift from project-specific purchases to repeatable financing platforms democratizes access to high-end compute. Startups, enterprises, and smaller cloud providers without the massive balance sheets of tech hyperscalers can now secure factory-scale hardware at institutional capital costs. By establishing compute as a stable, bankable asset, the initiative secures a continuous pipeline of funding to scale up AI development.