NVIDIA unveils $600B AI factory opportunity with up to 1.5M GPUs per site

NVIDIA is framing the AI factory as a massive business opportunity. As reported by 36Kr on August 18, the company unveiled a $600 billion market opportunity around AI factories, where a single factory can demand up to 1.5 million GPUs.

The framing comes from NVIDIA's latest infrastructure statement. In its official blog, the company describes AI factories as the defining infrastructure of the AI era — places where compute transforms energy and data into intelligence that powers every business, industry and country. In the AI economy, compute is revenue.

NVIDIA stresses that AI factories require a full stack of critical resources: advanced chips, packaging, memory and networking, as well as land, power and other supporting conditions. Building an AI factory is therefore not just a GPU purchase but an investment spanning semiconductors through energy infrastructure.

The announcement lands amid a dramatic surge in AI infrastructure spending, with cloud providers and compute operators racing to plan hyperscale data centers. By framing the market at the scale of 1.5 million GPUs per factory and a $600 billion opportunity, NVIDIA is also setting up demand for its next wave of chips, networking and interconnect gear.

For investors and the supply chain, the numbers push the ceiling on AI compute investment even higher — a single project could exceed the scale of an entire traditional data center, revaluing supporting areas such as land, power and cooling.

What to watch next: the actual pace at which these giant AI factory projects break ground, policy support for land and power resources, and how NVIDIA's next-generation GPUs and interconnect technology perform at such extreme cluster scale.