Jensen Huang says Nvidia can no longer design chips or write software without artificial intelligence, describing a dependence that already extends across two central parts of the company’s work.

“We can’t design a chip anymore without AI,” he said. His explanation centres on the range of designs the company can afford to explore: AI systems work overnight, examining possibilities that Nvidia would otherwise regard as too expensive to investigate.

The distinction matters. Huang is describing AI as a means of widening the search for possible designs, rather than simply completing an existing task more quickly. In his account, cost limits what the company would attempt without it.

He made an equally categorical statement about software: “We can’t write software without, without AI anymore.”

Huang also wants every part of Nvidia to use AI aggressively, extending its role beyond the chip-design and software work he described. His ambition is to turn the company into what he called “a one giant AI”.

That is a company-wide ambition, not a detailed account of how such an organisation would operate. His remarks establish the direction he wants Nvidia to take, but do not quantify the savings or explain how responsibility would be divided between people and AI systems.