NVIDIA CEO Jensen Huang recently shared a personal IPO-era story that puts the company’s meteoric rise into perspective. After NVIDIA went public, Huang sold some of his shares when the company was valued at around $300 million so he could buy his parents what he considered a truly special gift: a Mercedes S-Class. Looking back now—with NVIDIA valued at nearly $5 trillion—he called it his “only regret,” joking that it was the most expensive car in the world because of what those shares would be worth today.
Huang’s anecdote lands at a time when “I sold too early” stories have become a familiar theme around NVIDIA’s stock. Many investors have voiced frustration about missing the AI-driven run-up, and even high-profile executives have publicly acknowledged that selling past holdings in NVIDIA now looks painfully premature. Huang’s story stands out because it’s both relatable and symbolic: few people predicted just how central NVIDIA would become to the global AI boom.
Beyond the humor, Huang’s bigger message is about what’s driving NVIDIA’s surge in the first place. In a conversation at the World Economic Forum with BlackRock CEO Larry Fink, he argued that AI is fueling what he described as the largest infrastructure build-out in human history. According to Huang, the world is already “a few hundred billion dollars into it,” and the spending isn’t slowing down. His point is straightforward: modern AI models and the applications built on top of them require massive computing capacity, and that demand is forcing companies to build out entirely new layers of data-center infrastructure.
Huang also emphasized that this isn’t a small, incremental tech upgrade—it’s a multi-trillion-dollar shift. As businesses and hyperscalers race to deploy more powerful AI systems, they need the underlying compute and the end-to-end platform that makes AI training and inference possible at scale. That reality is a core reason NVIDIA has become so pivotal: its GPUs power a huge share of today’s AI workloads, and its broader platform helps developers and enterprises actually put that compute to work.
Still, the infrastructure boom comes with real questions. The biggest one is whether these enormous investments will translate into sustained, widespread AI adoption across industries over the long term. That uncertainty is also why interest in “agentic AI” has been climbing—companies are seeking AI systems that do more than respond to prompts, aiming instead for tools that can take actions, complete tasks, and deliver measurable business value.
Huang positions NVIDIA at the center of this transformation, not only because of its hardware, but also due to the surrounding ecosystem. He points to the company’s software and frameworks, including CUDA, along with its expanding lineup of AI technologies and models such as Nemotron. This mix of chips, software, and tooling has helped NVIDIA become more than a component supplier—it’s increasingly seen as a foundational platform for AI development and deployment.
In other words, the same force that made Huang’s Mercedes purchase an all-time expensive gift is the force reshaping global tech spending: the belief that AI is ushering in a new computing era, and that the infrastructure required to support it will be built at massive scale.
