Listen to the audio version of this article (generated by AI).

Hello, Reader.

If you think of the investment landscape as a complex puzzle, each piece needs to interlock to create the complete picture. Solving one section uncovers another.

Take this for instance: Three words investors dread: "supply is ____"

That would be "limited." And this term is rapidly gaining relevance in the context of AI investments.

The craving for AI solutions is unprecedented: companies are demanding more chips, servers, energy, and data centers. Yet, the supply of these critical resources can't keep pace with this insatiable appetite.

When supply falters, the firms that control access to these vital inputs are positioned to prosper significantly. This is why understanding the emerging bottlenecks in the AI sector is essential for investors.

In this installment of Smart Money, we’ll dissect the primary constraints affecting AI advancements, how these limitations could dictate which companies succeed, and the strategic avenues for your investments.

Identifying AI's Constraints

Let’s begin with a fundamental requirement for AI: energy. 

Data centers rely on high-performance chips from giants like Nvidia (NVDA) and AMD (AMD). However, these chips are ineffective without adequate power; without it, they essentially serve as expensive paperweights.

Therefore, while power is a necessity for AI expansion, it is also practically its backbone. Currently, demand for energy around data centers is overwhelming local power grids. In some regions, electricity prices have surged by as much as 267% in just five years. The next leaders in AI innovation won't just be those developing cutting-edge algorithms; they will also include the power providers fueling this vital infrastructure.

To satisfy this burgeoning demand, a multifaceted energy strategy is essential, incorporating wind, solar, nuclear, and natural gas sources. Industry titans like Microsoft (MSFT), Alphabet (GOOGL), and Amazon (AMZN) are already making substantial investments in diverse energy resources to secure reliable electricity for their future AI endeavors. For instance,

• Microsoft has entered a 20-year agreement for power from the anticipated revival of the Three Mile Island nuclear facility in Pennsylvania.
• Alphabet has teamed with Kairos Power to explore energy from small modular reactors (SMRs).
• Amazon is committing $20 billion to AI data centers in Pennsylvania, situating a campus close to the Susquehanna nuclear plant to ensure stable energy supply.

This scenario highlights that energy is becoming a strategic asset. But this isn't the only bottleneck we're grappling with; another critical component for AI systems is...

Memory: The Core of AI Functionality

Artificial intelligence also hinges on memory—specifically, DRAM.

A shortage of DRAM could stifle AI systems, limiting their capacity to process vast amounts of data. Currently, nearly 100 gigawatts of new data centers are projected to be operational within the next four years, but available DRAM will only support about 15 gigawatts in the immediate two years ahead.

Lack of adequate memory means AI can’t effectively operate. As Nvidia CEO Jensen Huang aptly noted, “The memory bottleneck is severe.”

Elon Musk echoed this concern in SpaceX's (SPCX) most recent earnings call, stating, “The limiting factor currently is memory.”

Industry leaders are sounding the alarm about these bottlenecks, and their implications for AI investment strategies bear heavy consideration. Yet, energy and memory are just the tip of the iceberg.

The Historical Context of Resource Shortages

Tech-induced shortages of critical resources are not new phenomena. The dot-com era serves as a salient example. The rapid proliferation of the internet during that time created unforeseen beneficiaries beyond the companies solely developing online platforms.

An acute demand for metals like copper, tantalum, and germanium arose, essential in fabricating the physical infrastructure of the digital age. Unfortunately, mining and refining capabilities didn’t scale quickly enough, resulting in a classic supply bottleneck. Investors who forestalled these limitations had the opportunity to realise extraordinary returns.

Between 1998 and 2001, I recommended four mining stocks to my audience that reaped impressive gains, emerging as unsung heroes during the tech boom.

One standout was Antofagasta plc (ANTO.L), a company pivoting towards copper mining. I flagged Antofagasta to my readers on December 18, 1998 – just before it commenced mining operations.

• Over the next three years, Antofagasta’s stock soared by 205%, while the S&P 500 languished around breakeven.
• In six years, Antofagasta generated a phenomenal 778% gain, while the S&P 500 dropped 27% during that period.

Antofagasta understood the resource constraints ahead and scaled its capacity throughout the late-1990s investment surge, reaping substantial rewards when the metal supply constricted.

This serves as a potent reminder of the value in spotting bottlenecks early on. Today, we encounter similar investment opportunities.

Decoding AI's Future Winners

The term limited is merely the initial insight into the AI investment puzzle. To truly grasp where the significant opportunities lie, investors need to tackle four more questions:

1. Where is demand outpacing supply?
2. Which firms are positioned at these bottlenecks?
3. Is it feasible to increase supply quickly?
4. Has the market identified these opportunities yet?

For deeper insights, check out my free Market Shock presentation, where I further explore AI's physical limitations, including energy and memory, alongside a third looming bottleneck that could influence the next generation of AI leaders.

I’ll also outline the types of companies likely to reap benefits from limitations, featuring 15 free stock picks that I believe are well-positioned to capitalise on these AI shortages.

Understanding the dynamics of AI's capabilities is critical when shaping your investment portfolio. Remember, within every challenging puzzle, the most valuable pieces often remain hidden. By pinpointing the bottlenecks hindering AI's progress, smart investors can discover the entities ready to thrive by overcoming these challenges.

Learn more about this today.

Best,

Eric Fry