SANTA CLARA, September 20 — Jensen Huang has put a number on the end of the world, and the number is zero. Nvidia’s chief executive says there is a “0% chance” that AI destroys the world by 2030, dismissing the doom warnings associated with Anthropic and arguing the technology should be developed as fast as possible. The claim has reignited the argument over AI risk — and, just as pointedly, over Nvidia itself.
The company sits at the centre of the boom, and its product line is being reshaped around it.
A graphics card rebuilt for AI
Take the RTX PRO 5500 Blackwell. It carries the same core count as the RTX 5090 but with 2.6x the VRAM. Coverage of the launch framed gaming as taking a backseat, with Nvidia overhauling the RTX 5090 for maximum AI margins while the RTX Pro 5500 delivers 2.6X VRAM at matching specs.
The mechanism is straightforward enough: more memory per chip lets larger models sit closer to the processor, which is what buyers running AI workloads pay for. Demand, meanwhile, is the question that decides everything else.
Nvidia has said it will double chip sales next year, and the supply chain has been identified as the place where that outcome gets decided.
Who is buying, and where
AWS and Nvidia have announced plans to deliver 2 million additional GPUs and next-generation infrastructure for agentic and physical AI. India’s Yotta has bet $12 billion on 80,000 Nvidia Rubin GPUs, with a 2026 timeframe cited. The customer base is broadening beyond the hyperscalers into life sciences.
Lilly, Roche and BMS are all building AI supercomputers, each doing so differently, according to R&D World. That matters because it moves Nvidia’s order book away from a handful of cloud giants and into research budgets.
Competition is not standing still. AIMultiple’s roundup of top 30+ AI chip makers places Nvidia alongside its rivals, and China’s homegrown AI chip push has accelerated, raising the question of whether it constitutes a real threat to the company. Nvidia’s own manufacturing is being reshaped by AI as well.
One video is being used to teach a robot to build Nvidia’s next chip, according to Manufacturing Digital. Britannica Money’s corporate profile covers the company’s history, its GPUs and its artificial intelligence business.
What to watch next is whether the supply chain can carry the doubling Nvidia has promised, and whether the 2030 deadline Huang has waved away keeps its hold on the debate.
