Everyone loves to talk about the AI giants — the chip designers, the hyperscalers, the software platforms hoovering up headlines and investment dollars. But tucked behind the curtain of the AI revolution is a component so fundamental, so unglamorous, and so utterly indispensable that without it, none of the magic happens. We're talking about memory chips. And right now, Micron Technology is sitting at the centre of one of the most interesting dynamics in the entire technology sector.

This week, we're pulling back the curtain on why memory — not just processing power — is quietly becoming one of the defining bottlenecks of the AI era, and what that means for how you think about the broader market landscape.

Here's a number worth sitting with: a single AI training run for a large language model can require hundreds of gigabytes of high-bandwidth memory, and inference (the process of actually running an AI model for users) demands fast, efficient memory access at every single query. Multiply that by billions of daily interactions across global AI platforms, and you start to see the scale of the problem.

The memory chip market has historically been brutally cyclical — boom and bust, oversupply and shortage, rinse and repeat. For years, Micron and its peers rode these waves, with investors bracing for the inevitable downturns. But the AI supercycle is introducing a new variable: structural, persistent demand for a specialised type of memory called High Bandwidth Memory, or HBM.

HBM is the premium product that sits directly alongside AI processors, passing data at extraordinary speeds. Micron, alongside South Korea's SK Hynix and Samsung, is one of only three companies in the world capable of manufacturing it at scale. That's not a coincidence — it's the result of decades of engineering investment and billions in capital expenditure. The barriers to entry are enormous.

Markets have taken notice. The conversation around AI infrastructure spending has expanded well beyond the obvious names, and memory suppliers are increasingly part of that conversation. Whether that attention is fully priced in, overdone, or still underappreciated is exactly the kind of question that makes markets interesting — and exactly the kind of question we won't pretend to answer for you.

If you've been around investing circles for a while, you've probably heard the old saying: during a gold rush, sell shovels. The idea is simple — when an industry is booming, the companies supplying the essential tools and infrastructure can sometimes offer a more durable business case than the companies directly chasing the gold.

In investing, this is often called the picks-and-shovels approach: rather than betting on which AI application "wins," you look at the foundational suppliers that every winner will need regardless of the outcome.

The appeal is intuitive. If five companies are racing to build the best AI model, and all five of them need vast quantities of high-bandwidth memory to do it, then the memory supplier has five potential customers instead of needing to pick the right horse. The risk isn't eliminated — suppliers face their own pressures around pricing, competition, and capacity — but the exposure is diversified across the ecosystem.

The important nuance? Picks-and-shovels companies are not automatically safer or better investments. They still face their own cycles, competition, and execution risks. During a downturn, demand from every customer can fall simultaneously. The framework is a lens for thinking, not a guarantee of returns.

When you're exploring any booming sector — AI, clean energy, biotech — it's worth asking: who supplies the suppliers? Who makes the tools that make the tools? These second- and third-order questions often reveal corners of the market that the loudest headlines miss.

Before you dive into any sector or theme, try mapping the supply chain from the bottom up. Ask yourself:

This isn't about finding a "secret" stock. It's about building a richer mental model of how industries actually work — so that when you read earnings reports, analyst commentary, or market news, you're seeing the full picture rather than just the headline act.

The AI story is still being written. But the infrastructure underpinning it — including the memory chips powering every model, every query, every breakthrough — is a fascinating and often overlooked chapter worth understanding.

Thanks for reading Wealth Runway. If you found this useful, share it with a friend who's trying to make sense of the AI hardware landscape without the jargon. We'll be back next week with another deep dive into the markets, the mechanisms, and the mental models that help you invest more thoughtfully.

Stay curious. Stay informed. Never stop asking the second question.

— The Wealth Runway Team

This newsletter is provided for educational and informational purposes only and does not constitute financial, investment, tax, or legal advice. Nothing in this publication should be construed as a recommendation to buy, sell, or hold any security or other financial instrument. Always conduct your own research and consult a licensed, regulated financial advisor before making any investment decision. Past performance is not indicative of future results.

/