Jensen Huang called it an “inflection point.” Investors called it a buying opportunity. On Thursday, Nvidia shares jumped nearly 9%, adding roughly $440 billion in market cap in a single session, one of the largest single-day value gains in stock market history for any company. That’s not noise. That’s a signal.
According to CNBC, the surge followed Nvidia’s Q2 earnings release, which sailed past analyst estimates and included forward guidance that stopped investor anxieties in their tracks. CFO Colette Kress projected 70% revenue growth for fiscal 2028. Huang was even bolder, saying actual demand is “much greater than 70%” but that supply, not orders, is the limiting factor. That framing matters. Nvidia isn’t demand-constrained. It’s production-constrained, which is a very different problem to have.
The broader chip sector caught the wave. Broadcom and Intel both rallied. Neocloud provider Nebius also moved higher. These sympathy plays reflect just how central Nvidia has become to the AI infrastructure story. When Nvidia beats, the whole supply chain breathes easier.
Huang’s commentary about who is buying GPUs deserves attention. A year ago, he noted, a single lab was driving the build-out. Now the buyer pool includes frontier AI labs scaling in parallel, enterprise customers, industrial AI deployments, and startups. Nvidia’s ACIE segment, covering AI clouds, industrial, and enterprise customers, hit $40.3 billion in quarterly revenue, up 138% year over year. That diversification is exactly what analysts have been asking for. Hyperscaler concentration was a real risk. It’s becoming less of one.
Still, there are genuine headwinds here. Taiwan Semiconductor, Nvidia’s primary manufacturer, continues to face supply chain pressure. Memory chips, critical components in Nvidia’s systems, are also in short supply. These are structural bottlenecks that even record demand can’t fix overnight. And then there’s the custom silicon threat. OpenAI, Google, Amazon, and others are all building their own AI chips. Analysts flagged this directly on Thursday as a risk to Nvidia’s near-monopoly on advanced AI accelerators. It won’t hurt Nvidia next quarter, but it’s a pressure that compounds over time.
The reported $12.9 billion acquisition of Hugging Face, first cited by The Information, adds another layer to the story. If that deal closes, Nvidia would own one of the most widely used platforms for open-source AI models, pushing it well beyond hardware into the model and software ecosystem. That’s a strategic move AMD, Intel, and other chip rivals simply can’t match right now.
For developers and founders evaluating AI infrastructure: Nvidia’s position at the top of the stack is not softening anytime soon. The supply crunch means access remains competitive, and pricing power is intact. Plan accordingly.
