For Nvidia, the first fiscal quarter of 2027 was a study in contrasts. Revenue hit a record $81.6 billion, with the data-center business alone surging 92 percent to $75.2 billion. Yet beneath those headline numbers, investors are grappling with a stark reality: the company’s exposure to China has become an expensive liability, and the answer lies in a sweeping strategic realignment that reaches from sovereign governments to private cloud operators.
The China toll is deeper than a single write-off
The US government now requires Nvidia to obtain a license before exporting its H20 chips to China. The immediate cost came in the form of a $4.5 billion writedown on inventory and supply commitments. But the forward-looking damage is even bigger: the quarterly guidance baked in an estimated $8 billion in lost H20 revenue. Efforts to keep a narrow channel open have so far failed. Since February 2026, US licenses have allowed small shipments of H200 chips to select Chinese customers, but not a single dollar of revenue has materialized from those approvals. Chinese authorities are blocking imports at the border and warning local companies to avoid US AI technology unless absolutely necessary. Nvidia estimates the Chinese AI accelerator market at nearly $50 billion, and losing access, the company warns, would materially weigh on its business while handing an advantage to overseas rivals.
Sovereign AI: a $30 billion hedge
That loss is largely priced into the stock, but what makes the current moment intriguing is the structural shift Nvidia is engineering in response. The concept is straightforward: help nation-states build their own AI infrastructure, complete with national data centers, homegrown models, and local computing capacity. For governments pursuing digital sovereignty, the pitch is compelling. For Nvidia, it insulates the business from over-reliance on any single market. The numbers show the bet is paying off. Sovereign AI generated $30 billion in revenue during fiscal 2026, more than triple the prior year’s level. Hyperscalers still account for roughly half of data-center sales, but the other half now comes from AI clouds, industrial customers, enterprises, and sovereign clients. The geographic footprint is expanding quickly. In Europe, France, Italy, Spain, and the UK are building national AI infrastructure that will collectively deploy more than 3,000 exaflops of Nvidia Blackwell computing power. In the UAE, Aleria is using 8,640 Nvidia Blackwell Ultra GPUs, with plans to scale to 16,000, and has secured early access to DGX Vera Rubin supercomputers.
Vultr and the inference era
At the same time, Nvidia is locking in major commercial wins that reinforce the shift from training to inference. Private cloud provider Vultr has selected Nvidia and Hewlett Packard Enterprise to expand its data centers, adopting the new GB300 systems. Financial terms were not disclosed, but the deal reflects a broader industry pivot: applied AI in day-to-day business is driving demand more than pure model training. Vultr CEO J.J. Kardwell noted that three years ago startups dominated demand; today it’s established companies rolling out customer-facing applications. Nvidia also unveiled its Vera CPU at a trade show in Las Vegas, a processor designed for autonomous AI agents. It handles tool calls and processes data in real time, with shipments starting in 2027.
Should investors sell immediately? Or is it worth buying Nvidia?
Dividend hike and shareholder meeting
For shareholders, the near-term calendar is marked by next Wednesday, June 24, when Nvidia holds its virtual annual general meeting. The agenda includes routine items like board elections, but the more significant corporate action has already been taken: Nvidia is raising its quarterly dividend from one US cent to 25 cents per share, with payments landing in late June.
Stock: constructive but cautious
At €182.44, Nvidia’s stock sits 12 percent above its 200-day average and just above the 50-day line. The RSI of 51.2 signals neither overbought nor oversold conditions. Of the analysts surveyed by S&P Global, 62 rate the stock a consensus “Strong Buy.” The roughly 4 percent decline over the past 30 days stands against a 12-month gain of nearly 44 percent. That captures the picture concisely: the long-term structural uptrend remains intact, while the short-term path runs through a geopolitical minefield no earnings model can fully capture. The gap to the 52-week high of €202.50 is not a technical problem—it is a measure of how much uncertainty the market is still pricing into the world’s most important AI infrastructure company.
The management has signaled that it expects industry investment to remain elevated, with major cloud providers projected to spend $1 trillion annually by 2027. Last quarter’s 85 percent revenue growth supports that outlook, and the upcoming shareholder meeting offers the next concrete opportunity for deeper insights.
Ad
Nvidia Stock: Buy or Sell?! New Nvidia Analysis from June 18 delivers the answer:
The latest Nvidia figures speak for themselves: Urgent action needed for Nvidia investors. Is it worth buying or should you sell? Find out what to do now in the current free analysis from June 18.
Nvidia: Buy or sell? Read more here...
