Nvidia's Record Quarter and $150 Billion Taiwan Vow: The AI Giant's Balancing Act Between Supply and Expectations

Despite an 85% revenue surge and massive shareholder returns, Nvidia's stock barely moves. Optical networking bottlenecks and $150B Taiwan expansion plans dominate outlook.

Nvidia just delivered another blowout quarter, raised its dividend twenty-five-fold, and unveiled an $80 billion buyback — yet its shares barely budged. The stock edged lower to around $210 following the earnings release, underscoring how extraordinary performance has become the baseline for a company that has redefined the AI landscape. On a euro basis, the shares currently trade at €185.36, roughly 8% below the 52-week high set on May 14, though still 15% above the 200-day moving average. The RSI of 36.4 signals mild oversold conditions after months of consolidation between $210 and $230.

The revenue numbers are staggering by any measure. For the first quarter of fiscal 2027, which ended April 26, 2026, Nvidia posted $81.6 billion in sales, an 85% year-over-year surge. The data center segment alone contributed $75.2 billion, up 92% from the prior year. Management credits the relentless shift toward "agentic AI" — systems capable of executing tasks autonomously — for sustaining demand. The company guided for roughly $91 billion in second-quarter revenue, a trajectory supported not only by hyperscalers but increasingly by smaller enterprise customers. On the profitability front, Nvidia maintained a non-GAAP gross margin of around 75%, though analysts flagged rising memory prices as a potential headwind.

That insatiable demand is colliding with a critical bottleneck: optical networking. According to a research report from Rosenblatt Securities, Nvidia has asked its suppliers to boost capacity for indium phosphide lasers by up to twenty-fold by 2030. The component is central to co-packaged optics, a technology that sits next to the chips and promises higher data-transfer speeds with lower power consumption — essential for stitching together clusters of thousands of GPUs. Suppliers themselves consider a twelve-fold increase more realistic, underscoring the tension between ambition and execution. The message from Santa Clara is clear: optical links are now the narrowest passage in the AI supply chain.

The answer, in part, lies in a massive expansion of manufacturing partnerships on the island. CEO Jensen Huang, speaking at the TD Cowen conference on May 28, disclosed that Nvidia now plans to spend between $100 billion and $150 billion annually with fabrication partners in Taiwan — up from a previous range of $10 billion to $15 billion. Describing Taiwan as the "epicenter of the AI revolution," Huang said the funds will underpin production of the Blackwell Ultra architecture and the next-generation Vera Rubin platform, with first systems set for delivery later this year. The next big reveal is slated for June 1 at the GTC Taipei event on Computex, where Huang is expected to detail the Vera Rubin roadmap and possibly introduce new products for physical AI and robotics.

Navigating geopolitics adds another layer of complexity. Huang simultaneously joined the advisory board of Tsinghua University in Beijing, a panel that includes Apple's Tim Cook and Tesla's Elon Musk. The move aims to preserve communication channels to the Chinese market, where Nvidia recently restructured its operations — but it has already drawn scrutiny from Washington. It is a delicate balance for a company that derives a growing share of its revenue from data-center chips while trying to keep access to the world's second-largest economy.

The board’s decision to raise the quarterly dividend to $0.25 per share — a twenty-five-fold increase from the previous $0.01 — and authorize an $80 billion buyback program underscores the confidence in cash generation. Yet the market's muted response reflects a harsh reality: Nvidia's performance is now priced for perfection, and even record-breaking results struggle to spark a re-rating higher. With the Vera Rubin launch and the GTC keynote on the horizon, the next leg of the story depends on whether execution can outrun already sky-high expectations.