AI infrastructure, not algorithms, emerging as key AI growth constraint: Nvidia's Jensen Huang

New Delhi [India], August 18 (ANI): Nvidia CEO Jensen Huang has sought to underline the growing importance of physical infrastructure for the next phase of artificial intelligence, arguing that access to land, power and data-centre shells is increasingly becoming as critical as advanced chips and computing systems for building AI factories.
In a post on X, Huang said the AI industry was entering a phase where infrastructure availability could determine how quickly companies expand their AI capabilities, particularly as frontier AI laboratories see demand for computing capacity grow faster than their ability to independently secure the underlying infrastructure.
"AI factories are the defining infrastructure of the AI era," Huang said, describing facilities where computing transforms energy and data into intelligence for businesses, industries and countries.
Huang's post focused on Nvidia's arrangement with SB Energy to secure land, power and shell (LPS) capacity at the PORTS-Pike Technology Campus in Portsmouth, Ohio, where OpenAI is expected to be the tenant.
The Nvidia CEO said that while large cloud service providers and investment-grade enterprises generally have the financial strength and infrastructure expertise to secure such resources themselves, frontier AI labs face a different situation.
"Their growth is increasingly constrained not by algorithms or customer demand, but by the availability of compute," Huang said.
The post highlights Nvidia's strategy of moving beyond supplying GPUs and other computing components to securing some of the physical infrastructure needed to deploy its systems. Huang said Nvidia would be selective in doing so, focusing on sites where customer demand is sufficiently visible and durable to support multiple generations of computing systems.
The PORTS-Pike facility is expected to initially provide 4.25 gigawatts of AI factory capacity. Nvidia said each generation of systems deployed at the site could represent approximately 1.5 million Nvidia GPUs, translating into an estimated USD 150-200 billion in Nvidia revenue per generation.
Huang also sought to address concerns over Nvidia's financial support for the infrastructure, stressing that the arrangement was not circular financing.
"No. OpenAI will pay the lease," he said, adding that Nvidia was using its scale and visibility into customer demand to secure critical infrastructure that would host Nvidia computing capacity.
AI is increasingly becoming an infrastructure-intensive industry, with power availability, land, financing and data-centre capacity emerging as strategic resources alongside semiconductors.
Huang said Nvidia's role has evolved from building accelerated computing chips to developing complete systems, networking, CUDA software and AI factories, and now to helping secure the physical infrastructure required to deploy those systems. (ANI)