At about 4:30 p.m., there was a moment. On May 20, traders watching their terminals nearly couldn’t believe what they were reading when they saw the numbers come across the wire. In its first fiscal quarter, Nvidia reported $81.6 billion in revenue, an 85% increase over the same period last year. Profit reached $58.3 billion, more than tripling. The earnings per share of $1.87 easily exceeded the forecast. Nevertheless, the stock hardly moved by the following morning. Some screens even had red flashes.
Two years ago, this kind of thing would have seemed unimaginable. At the time, a beat by Nvidia sent the entire semiconductor index into a frenzy, caused Asian markets to rise overnight, and led TV hosts to refer to it as “a moment for the history books.” Even a record-breaking quarter feels somewhat expected these days because the bar is so high. Perhaps that’s the price of becoming the most valuable company in the world.
You can feel the change as you stroll through Santa Clara. The Voyager and Endeavor buildings, with their tumbling glass panels, feel more like the architecture of a silent empire than offices due to the recent expansion of the Nvidia campus. According to reports, engineers are working on Blackwell’s next version and future iterations. The most recent report’s numbers explain why. This quarter, data center revenue alone generated $75.2 billion, a 92% increase. That’s more than a press release line. For some mid-sized national economies, that is almost all of their yearly output.
However, investors are beginning to pose more challenging queries. The number that analysts keep returning to is the gross margin, which stands at an astounding 74.9%. For a hardware company, it is practically unheard of. Even at the height of iPhone dominance, Apple remained far below that. Some on Wall Street believe that this cannot continue indefinitely. Alphabet, Amazon, and Meta’s custom chips are subtly eroding the edges. Within a few years, inference workloads are expected to dwarf training markets, which could eventually favor application-specific accelerators over general-purpose GPUs. Nvidia might continue to lead. Additionally, it’s possible that the margins shrink before anyone notices.
Perhaps more intriguing is what Jensen Huang said at a launch event for Nvidia’s new Taiwan headquarters last week in Taipei, almost casually. He informed reporters that the company now intends to increase its annual expenditure on its Taiwan supply chain from about $100 billion to about $150 billion. That amount was more like $15 billion five years ago. The trajectory is difficult to ignore. The geopolitics of silicon are being reshaped by the man in the black leather jacket who used to sell gaming graphics cards to PC enthusiasts in suburban computer stores.
Everything is still clouded by the China question. Nvidia held a nearly 95% market share for AI chips in China a year ago. It’s practically zero today. A complex web of export regulations, political negotiations, and the rate at which Huawei and other domestic players close the gap will determine whether that loss is temporary or permanent. While some investors think it’s already priced in, others see it as a wound. As of yet, there is no definitive solution.
Then there is the more general question of artificial intelligence. Recently, Gavin Baker, a fund manager overseeing over $4 billion, pointed out that the valuations of all the AI plays “just can’t all be accurate.” Nvidia and memory increase if the power, cooling, and optical names are correct. Many of the others perform poorly if Nvidia is correct. It’s an odd equilibrium that most likely won’t last forever.
As you watch this develop, it seems like we’re at one of those infrequent turning points that the market doesn’t fully realize until much later. By any historical comparison, the quarterly figures are remarkable. The price target set by Tigress Financial is $425. The business approved buybacks totaling billions more. Money is coming in more quickly than it can be reinvested. After the print, the stock continued to decline.
Perhaps the most telling detail is that. Greatness ceases to feel great when it becomes routine. Whether Nvidia can continue to expand is not the true question. The question is whether the market can continue to be taken by surprise.
