Nvidia's Berlin GTC ends with containerized AI data-center pitch as neocloud stocks slide

Huang met with world leaders including U.K. Prime Minister Andy Burnham, French President Emmanuel Macron and German Chancellor Friedrich Merz during the trip, announced new products and appeared on stage with small robot ducks. At a media Q&A, Huang described his approach to bringing data centers online faster: 'Try less sleep, which is what I do.' The trip continued with a stop in Kazakhstan, where Huang was touting the country's potential as a major AI hub, CNBC reported.

On stage in Berlin, Huang announced a product that packs seven of Nvidia's Vera Rubin AI racks into a shipping container and is designed to be moved on normal roads. The pitch is to bring AI compute online faster as demand surges. Huang said delays in getting data centers running can cost billions of dollars in revenue, and that moving the system into the supply chain would speed up when computing capacity can be turned on.

The final day of the conference came as public neocloud stocks sold off. CoreWeave ended the day down 10% and Nebius dropped 7% by market close, continuing a turbulent period for companies that lease AI compute. Many of their shares have fallen on concerns about heavy borrowing, CNBC reported.

Huang, whose company holds sizable stakes in many of those businesses, remained bullish. 'If you look at the fundamentals, the demand for the neo clouds is off the charts,' he told CNBC.

At GTC, other attendees were more cautious. One executive at an AI infrastructure company said the 'magnitude of businesses with no credit rating' was 'freaking the market out,' according to CNBC.

AI infrastructure is expensive to build. Goldman Sachs estimates AI-related capital spending could total about $7.6 trillion from 2026 through 2031 across compute, data centers and power, CNBC reported. Huang told the Berlin audience that a 1 gigawatt data center costs about $50 billion. In an interview with CNBC, he said working capital to fuel growth is missing because demand across AI sectors is so large. 'The demand of the capital is much greater than the availability of the capital,' he said, adding that the issue will be solved. An executive at an AI infrastructure financing company told CNBC on the sidelines that 'very soon capital will become the bottleneck.'

Major compute consumers are already exploring smaller data center deployments. Sources told CNBC in September that OpenAI and Anthropic had been hunting for smaller AI data center deals.