The Infrastructure for the AI Revolution

Erik@YWR on NVIDIA Corporation (NVDA-USA | nvidiacorpor)

5/31/2025

Summary

Nvidia has transitioned from being the leading GPU company to becoming the cornerstone of the artificial intelligence (AI) infrastructure ecosystem. What was once a company associated with gaming graphics is now the indispensable enabler of AI computation, software, and data manufacturing—what its visionary CEO, Jensen Huang, now calls “AI Factories.” Despite its meteoric stock price rise, Nvidia remains underappreciated by many institutional investors. While some dismiss it as a crowded trade or an overvalued momentum stock, the company is deeply embedded in the future of compute, robotics, and industrial transformation. With the AI revolution still in its early innings, Nvidia stands out not only as a technology supplier but as a systems orchestrator, defining entire categories and standards through CUDA, Blackwell chips, and vertically integrated solutions. This investment view argues that Nvidia should not be seen through the traditional lens of semiconductor cycles or dot-com analogies. Instead, it should be seen as the Microsoft of the AI age—critical infrastructure with enormous pricing power, growth prospects, and network effects. Backing Nvidia is not simply buying a chip stock—it’s backing the Jockey (Huang) who is shaping the next industrial age. By 2027 NVDA could be trading over $200/share as investors look ahead to a potential 2029 EPS of over $9/share and put a 25x multiple on those earnings.

Thesis

1. Jensen Huang: One of the Greatest Founders of All Time The Nvidia thesis begins with its founder. Jensen Huang is emerging as one of the most consequential entrepreneurs of our era, standing shoulder to shoulder with Jobs, Bezos, Musk, and Gates. His long-term thinking, relentless execution, and ability to articulate a vision that unites hardware, software, and manufacturing partners makes him unique. While competitors remain reactive and defensive, Huang is building entire ecosystems—from CUDA as the foundational AI programming language to integrated hardware-software pipelines like Omniverse for simulation. He’s also charismatic, healthy, tireless, and strategic. The 100-minute keynote he delivered at Computex 2024 was not just a product showcase; it was a roadmap for where AI and computation are headed. It was also a message to institutional investors: “We’re building AI factories, not datacenters. Get onboard.” Backing Nvidia means backing a founder who is redefining the rules of the game and playing at an entirely different tempo. 2. Nvidia Is Not a Chip Company. It’s the AI Infrastructure Layer. The perception of Nvidia as a “chip stock” is outdated. Nvidia is now building the infrastructure layer of the AI era. From GPUs to full-stack systems to simulation platforms like Omniverse, it is becoming the compute layer, the training layer, and the execution layer of artificial intelligence. What used to be datacenters are now “AI factories.” These facilities don’t store data—they manufacture intelligence in the form of tokens. As Google’s own metrics showed, they scaled from 9.7 trillion tokens to 480 trillion/month in just a year. Nvidia is supplying the machines for this next generation of compute. The implications are profound. Not only is Nvidia embedded in training LLMs, it is also powering inference, robotics, digital twins, and the core infrastructure for any enterprise seeking AI capabilities. From AWS to Meta to startups, Nvidia is the default option. 3. The AI Wave Is Just Getting Started The AI boom is still in its infancy. While ChatGPT brought large language models into the mainstream, the broader AI integration into business processes, logistics, robotics, and simulation is just beginning. Nvidia’s Computex keynote showed how we’re still early in the rollout of AI-native enterprise software, robotic automation, and “containerized thought.” Nvidia is not just riding this wave—it’s creating it. Its early bets on software-defined infrastructure, its investment in frameworks like CUDA, and its end-to-end vertical solutions ensure that it is not just a supplier but the default platform for AI. Every wave of innovation—from cloud to AI to robotics—builds on Nvidia’s tools. This compounding advantage is one of the strongest moats in tech today. 4. Taiwan Is the Nexus—and Nvidia Owns It An underappreciated part of the Nvidia story is its deep integration into Taiwan’s semiconductor and manufacturing ecosystem. Companies like Wistron, Pegatron, Wiwynn, and Quanta—all key Nvidia partners—are leading the AI hardware transformation. Jensen’s tribute videos to Taiwan’s AI supply chain showed how Nvidia has pre-locked its manufacturing pipeline through close relationships with TSMC and others. With Taiwan as the epicenter of software-defined manufacturing and robotic factories, Nvidia is tapping into a unique, irreplaceable infrastructure advantage. This also explains Nvidia’s urgency in seeking a “Plan B” with Samsung’s foundry. The AI infrastructure race is global, but Nvidia’s position at the heart of this Taiwanese ecosystem is unmatched—and hard to replicate. 5. Deepening Moats in Software and Storage While competitors scramble to build chips, Nvidia is redefining the software and storage layers to suit its architecture. CUDA is now the dominant programming layer for AI, and competitors have largely conceded this space. In storage, Nvidia’s emphasis on reinventing the enterprise tech stack is a massive tailwind for companies like NetApp (NTAP), which are already being integrated into Nvidia’s AI infrastructure vision. A generational replacement cycle in enterprise storage and compute is now underway, with Nvidia orchestrating it. Owning Nvidia gives you access to this broader transformation across tech infrastructure. 6. This Is Not Cisco 1999 The fear that Nvidia is the next Cisco—overhyped in the dot-com era and underwhelming thereafter—is misplaced. Jensen is deploying a multi-pronged strategy across verticals: chips, enterprise, robotics, simulation, cloud, consumer. Nvidia is becoming indispensable not just to startups but to governments, factories, banks, and healthcare systems. His famous quote—“I would love if you would buy everything from Nvidia, but at least buy something from Nvidia”—shows the breadth of the strategy. Nvidia doesn’t need to dominate every vertical. It just needs to be present everywhere. That’s how platforms win. 7. Valuation Is Surprisingly Reasonable Despite its stock price and dominance, Nvidia trades at roughly 30x forward earnings—one of its lowest multiples historically. When you compare that to other category-defining tech companies in their early days, it looks cheap. If Nvidia can grow EPS to $9.20 by 2029, and you apply a conservative 30x multiple, you get a year-end 2027 price target of $200/share. This excludes any upside from buybacks (currently small) or dividend expansion. In fact, given its rapidly growing free cash flow, Nvidia has multiple levers to create shareholder value beyond organic growth. For a company at the center of the biggest tech wave in decades, this valuation seems more than fair—it seems like an opportunity.

Risks

1. Geopolitical Risk in Taiwan Nvidia’s supply chain is deeply tied to Taiwan. Any disruption due to geopolitical tensions—particularly around China-Taiwan-US relations—could impact production. While Nvidia is actively diversifying via Samsung and U.S.-based fabs, this risk is real and non-trivial. However, the U.S. has a vested interest in maintaining Taiwan’s stability, not least because of Nvidia’s strategic importance. Still, investors must factor this into the risk-reward calculus. 2. Hype Cycle Reversal or Air Pockets There is always the risk that AI proves slower to commercialize than expected, or that massive capital expenditures by hyperscalers (e.g., for $50 billion AI factories) face delays, cost overruns, or shifts in demand. In such scenarios, Nvidia’s growth could face temporary slowdowns. While the structural thesis remains intact, stock price volatility could be significant. 3. Competition in Chips and Platforms Companies like AMD and Intel are not standing still. Microsoft and Google are developing custom silicon (Azure Maia, Google TPU). The risk is that hyperscalers vertically integrate, reducing reliance on Nvidia’s hardware. However, Nvidia’s software ecosystem (CUDA), scale, and pace of innovation make it hard to displace. Still, this is a space to watch. 4. Cyclical Nature of Capex Spending Despite its growth, Nvidia is still partially exposed to the capex cycles of large technology buyers. A pullback in tech or economic slowdown could impact near-term revenues, even if the long-term thesis holds. 5. Founder Risk Much of Nvidia’s strategic magic comes from Jensen Huang himself. While he shows no signs of slowing down, any sudden departure would be a major risk. The depth of Nvidia’s bench is strong, but replacing a founder of this caliber is never easy.

📈 Price Targets

Tags