NVIDIA has become the essential infrastructure layer of the AI boom, and a major Singapore investment fund just bet billions on that thesis. Temasek Holdings announced it is doubling its artificial intelligence exposure from 6% to as much as 15% by March 31, 2031, deploying capital across five areas of the AI value chain. The move signals that smart investors are following the enablers of AI, not just the companies racing to build the most powerful models.

What Makes NVIDIA the "Picks and Shovels" of AI?

NVIDIA's position in the AI ecosystem mirrors the gold rush analogy perfectly. During the California Gold Rush, the companies that sold picks and shovels to miners often made more money than the miners themselves. NVIDIA plays that role today. The company provides the graphics processing units (GPUs), or specialized chips, that every major AI lab needs to train and deploy large language models, the AI systems that power tools like ChatGPT and Claude.

But NVIDIA's moat runs deeper than just selling hardware. The company built CUDA, a proprietary software platform that powers every major cloud and edge device. This creates what economists call "switching costs," meaning that once a company commits to NVIDIA's ecosystem, moving to a competitor becomes prohibitively expensive. That stickiness translates to predictable, long-term revenue streams regardless of which AI company wins the race for the most advanced models.

The numbers back this up. In the first quarter ended April 26, 2026, NVIDIA's revenue surged 85% year over year to a record $81.6 billion. The Data Centre segment, which serves AI companies and cloud providers, grew even faster at 92% to $75.2 billion, driven by massive ramp-up of the new Blackwell architecture. Blackwell represents NVIDIA's latest generation of chips, designed specifically for the computational demands of training and running today's largest AI models.

"The extraordinary speed of the buildout of AI factories," explained CEO Jensen Huang, describing the demand driving NVIDIA's growth.

Jensen Huang, CEO at NVIDIA

This growth matters because it shows that NVIDIA benefits from competition, not from any single winner. Whether OpenAI, Alphabet, or another frontier AI lab dominates the race to build the most capable models, they all need NVIDIA's chips. The company has already positioned itself to profit from the entire AI gold rush.

How to Identify AI Infrastructure Winners in Your Portfolio?

• Look for Switching Costs: Companies that create high costs for customers to switch to competitors tend to have more stable, predictable revenue. NVIDIA's CUDA platform and Microsoft's enterprise software integration create this effect.
• Follow the Enablers, Not Just the Winners: Instead of betting on which AI company will dominate, invest in the companies that serve all of them. Temasek is deploying capital across five areas: energy and data centres, semiconductors, cloud services providers, foundation models, and AI applications and software infrastructure.
• Prioritize Vertically Integrated Mega-Caps: Companies with expertise spanning multiple parts of the AI value chain, like Microsoft, have more resilience and more ways to capture value than single-focus competitors.
• Examine Real-World Productivity Gains: The most sustainable AI investments are those backed by genuine productivity improvements that justify the cost, not just hype or speculative bets.

Why Taiwan's Chip Monopoly Matters for AI?

NVIDIA designs the chips, but Taiwan Semiconductor Manufacturing Company (TSMC) manufactures them. TSMC controls more than 90% of global advanced chip manufacturing, making Taiwan a critical chokepoint in the global AI supply chain. In the second quarter of 2026, TSMC's revenue increased 33.7% year over year to $40.2 billion, with net profit climbing 77.4% to NT$706.56 billion.

Like NVIDIA, TSMC's business remains resilient regardless of which AI company wins. After spending decades mastering the complexities of running advanced fabrication plants, TSMC is currently the only foundry mass-producing the leading-edge AI chips designed by NVIDIA and other major tech companies. This monopoly position means TSMC, like NVIDIA, profits from the entire AI ecosystem rather than betting on any single winner.

What Role Does Microsoft Play in the AI Infrastructure Race?

Microsoft represents a different kind of AI infrastructure play. While NVIDIA and TSMC provide the hardware layer, Microsoft is building the software and cloud ecosystem layer. The company has already established itself as a leading cloud computing provider through Azure, secured a multi-billion-dollar stake in OpenAI, and embedded Copilot AI into its ubiquitous commercial software suite.

For the fourth quarter of fiscal year 2026 ended June 30, 2026, Microsoft's revenue increased 18% year over year to $90 billion, while diluted earnings per share rose 32% to $4.81. Azure and other cloud services revenue grew 43%, while paid Microsoft 365 Copilot seats passed 30 million, with net seat additions more than doubling quarter over quarter.

Microsoft's strategy creates enterprise stickiness similar to NVIDIA's switching costs. Once Copilot is wired into enterprise users' daily workflows, the cost of switching to another provider becomes extremely high. Microsoft is funnelling its operating cash flow into infrastructure, with capital expenditure expected to reach roughly $175 billion in calendar year 2026, a figure revised down from about $190 billion because of an accounting change, not because the company is spending less.

Why Temasek's AI Bet Signals a Shift in Investment Strategy?

Temasek's decision to double its AI exposure reflects a fundamental insight about how value flows in the AI economy. Rather than betting on which frontier AI lab will build the most capable model, the Singapore investment fund is betting on the companies that enable all of them. Temasek CEO Dilhan Pillay explained that the company's investments reflect the view of AI as a structural, long-term driver of value creation.

This approach acknowledges a hard truth: predicting the winner of the AI race is nearly impossible. But predicting that the race will happen, and that it will require enormous amounts of computing power, specialized chips, cloud infrastructure, and software platforms, is much easier. The companies providing those tools have already proven they can scale, generate cash flow, and maintain competitive advantages that are difficult to disrupt.

Smart investors recognize that while AI-enabling tech titans and proven adopters are expected to be key beneficiaries of AI, it is crucial to lean towards those supported by real-world productivity gains rather than pure speculation. The companies profiting most from the AI boom may not be the ones making the headlines with the latest breakthroughs, but rather the ones quietly powering the infrastructure that makes those breakthroughs possible.