Nvidia's AI Map Expands From Kazakhstan to the Phone in Your Pocket

Nvidia chief executive Jensen Huang has floated Kazakhstan as a potentially significant site for AI computing capacity serving Europe, pointing to the country's energy resources and its data-center buildout plans, CNBC reported. The remark lands alongside a burst of Nvidia announcements that stretch from a live mobile network to a processor switch at a cloud startup — and against a Wall Street backdrop that grew noticeably more cautious about how the AI buildout gets paid for.

Huang framed the Kazakh opportunity around two conditions: available power and expanding data centers. His comment described a possible role for the country beyond its home market, with the computing capacity potentially serving European users as well. He was explicit that this is a possibility, not a settled development — a distinction that matters when weighing the stock, since a geographic prospect carries a different signal than signed customer orders.

ByteDance, Alibaba and Tencent Place a $35 Billion Bet

How large demand for hardware can already run was illustrated by a separate report from China. Reuters reported Tuesday that ByteDance, Alibaba and Tencent had together ordered more than 3 million RTX PRO 5500 graphics cards, with a combined list price above $35 billion. Orders from Chinese customers overall covered more than 6 million cards.

Nvidia, however, warned that it cannot guarantee it will fill those orders. Interest alone does not determine which shipments actually become possible, so investors have to watch execution along the supply chain as closely as the appetite for additional computing power.

A $1.1 Billion Reservation at King Yuan Electronics

Capacity is being addressed upstream as well. On Wednesday Nvidia announced a $1.1 billion prepayment to reserve capacity at King Yuan Electronics, earmarked for new test facilities and equipment in the United States. Nvidia expects total semiconductor test capacity to double between 2026 and 2027. That expansion forms the industrial backdrop to Huang's site considerations: while Kazakhstan could host additional AI data centers, the reservation tackles an earlier step in production.

T-Mobile, Nvidia and Nokia Complete First AI-RAN Call

On the application side, Nvidia on Wednesday reported a milestone in mobile networks. T-Mobile, working with Nvidia and Nokia, completed what the companies describe as the world's first AI-RAN call on a commercial network. Unlike a plain product launch, the announcement covers a completed deployment in a live commercial network — a concrete use case for Nvidia technology outside classic AI compute provisioning.

BT, Deutsche Telekom and Elisa, meanwhile, announced tests or pilot projects that will run Nokia software on Nvidia's Aerial RAN Computer. Those efforts remain distinct from the completed call: they describe the next round of trials, not widespread operation.

Modal Switches to Vera; Germany Deal Targets Physical AI

Additional ground is opening up in AI agents. Reuters reported Monday that AI cloud startup Modal plans to use Nvidia's Vera processor for its operations, having previously relied on chips from Intel and AMD. For investors, the significance lies in the planned switch of technology — the news concerns not just added compute capacity but the choice of processor supplier for a specific task.

The same day, Nvidia and the German government agreed to cooperate on models for "physical AI" and on industrial AI applications. According to Nvidia, smaller and mid-sized companies as well as research institutions are to gain access to the models.

Applications Don't Answer the Financing Question

Together, the announcements trace several routes by which Nvidia aims to push its technology into more enterprise applications, though they sit at very different stages of maturity: a completed mobile call stands beside pilot projects, a planned processor deployment and a cooperation agreement.

That distinction is central to how the stock is read. Concrete applications demonstrate progress in usage, but they do not by themselves resolve the question raised Thursday, when an AI stock selloff — described by the Associated Press as driven by worries over rising credit costs for data centers and chip purchases, plus doubts about whether valuations had run ahead of business fundamentals — weighed on trading. Nvidia's operational news and investors' more guarded stance are not contradictory; they concern different preconditions for further growth.