Jensen Huang’s Senate No-Show and the Quiet March of Nvidia’s AI Empire

Nvidia is living two lives at once. In Washington, the chipmaker is sparring with Senator Elizabeth Warren over a subpoenaed hearing on export controls. In the corporate world, it is stitching together new alliances with Apple, Deloitte, and the data-center giants that will power the next wave of artificial intelligence. That tension — regulatory friction on one side, accelerating infrastructure demand on the other — defines the stock’s current moment.

A summons ignored

Senator Warren had invited Jensen Huang to testify before the Senate on June 11, focusing on AI policy and restrictions on chip sales to China. Huang declined, offering instead a meeting at Nvidia’s headquarters. Warren hit back publicly, arguing that a CEO with time for expensive dinners and trips to Beijing should also make time for Congress. The dispute underscores how deeply Nvidia is entangled in the U.S.-China tech war.

The impact on the top line has been brutal. In a recent quarter, export curbs stripped away roughly $2.5 billion in revenue, followed by a billion-dollar writedown on inventory. Nvidia now models zero revenue from China in its forward guidance, while the Commerce Department continues to close loopholes that might allow shipments to bypass the rules. The Senate will debate new export regulations on June 11 — without Huang in the room.

Apple reopens the door

Away from the political fray, Nvidia notched a quiet but significant win. Apple will use Blackwell graphics processors in its new cloud infrastructure, running on Google Cloud. It is the first time in years that Apple has turned to Nvidia for data-center silicon. The system is expected to be fully operational by late summer 2026. The deal proves that, despite the regulatory noise, hyperscaler demand for Nvidia’s hardware remains voracious.

From chip peddler to platform builder

That demand is no longer just about raw GPUs. Nvidia is systematically transforming itself into an infrastructure company — and its London Tech Week announcement on June 8 illustrates the shift. Together with Deloitte, it launched the “Adopt 100” program, which vets AI solutions from startups and plugs them directly into Deloitte’s enterprise consulting practice. The goal: accelerate AI adoption inside big companies while embedding Nvidia technology deep into their workflows.

Should investors sell immediately? Or is it worth buying Nvidia?

The broader product line supports that ambition. At Computex 2026, Nvidia unveiled the Vera CPU, purpose-built for agentic AI and reinforcement learning, and the NemoClaw orchestration framework that gives firms blueprints for AI agents. The Nemotron 3 Ultra language model, with 500 billion parameters, was released as an open-weights model. The message is clear: Nvidia sells the entire AI stack, not just the chips. Its CUDA ecosystem, built over years, creates switching costs that competitors struggle to match.

Rubin heats up

The next-generation Rubin system is already in full production and will begin deployment at partner data centers in the second half of 2026. It pairs the Rubin GPU with the Vera CPU inside megawatt-scale racks designed for liquid cooling. These racks represent a new class of infrastructure — higher power density, custom thermal management, and tighter network integration. Nvidia wants to own that architecture, from the silicon to the rack to the software that ties everything together.

The market’s mixed verdict

The stock tells a more cautious story. On Wednesday, Nvidia shares traded at €177.00, down nearly 2% on the day and testing the 50-day moving average. From the record high set in May, the stock has shed roughly 12%. That slide comes despite a stellar fiscal year that saw revenue climb to nearly $216 billion. The political overhang — the risk of additional restrictions that could shrink Nvidia’s global market share — weighs heavily.

Still, analysts see room to run. The consensus price target stands at €257.88. The Wall Street Journal named Nvidia the most future-proof company in its 2026 ranking. And the company pays a quarterly dividend of $0.25 per share, with the ex-dividend date falling on June 4.

Structural logic

The real case for Nvidia is not about a single quarter or a single political spat. It is about controlling the infrastructure on which almost all AI progress runs — from language models to agentic systems to enterprise deployments. Every new partnership, every new platform like Rubin or NemoClaw, locks customers deeper into that ecosystem. The political headwinds are real and they may intensify. But the technological offensive is moving faster.