RJ Hamster
The Invisible Foundation of the AI Boom
The Invisible Foundation of the AI Boom
BY KEITH KAPLAN
CEO, TRADESMITH
Inside a factory the size of four football fields in Tainan, Taiwan, the air is cleaner than anywhere else on Earth.
Workers move carefully between the machines in what are known as “bunny suits” – white, head-to-toe coverings made from non-shedding polyester.
The suits aren’t to protect the workers. They’re to protect what the workers are making.
This is Fab 18 – one of the most advanced industrial plants on Earth. And it’s run by Taiwan Semiconductor Manufacturing Company, or TSMC, the world’s largest chipmaker.
Using beams of light, the machines inside print circuit patterns onto silicon wafers – patterns so small that 10,000 of them laid side by side would barely span the width of a human hair.
At these scales – measured in billionths of a meter – an out-of-place particle that would be invisible under a microscope can ruin a chip worth thousands of dollars.
What makes these mind-bendingly small scales possible is nanotechnology – atomic-scale engineering that lets chipmakers pack billions of transistors onto a silicon wafer the size of your fingernail.
It’s one of the investment themes lighting up on our new Global Trends Tracker, too, which we recently rolled out for our Platinum members. And it’s easy to see why…
Without the ability to build transistors just a few atoms wide – and pack billions of them onto a single chip – today’s AI models couldn’t run. That makes nanotech the invisible foundation of the AI boom.
So today, let’s look at what nanotechnology is… why it’s the invisible foundation of the AI boom… and how to profit from this critical investment theme.
Recommended Link
Don’t let another big opportunity pass you by…
If you missed Bitcoin in 2011, Tesla in 2012, or NVIDIA in 2016, you know the sting of regret… Now the master trader Jeff Clark has uncovered a new opportunity. He calls it “Crossfire” – and it could hand you gains of as much as 1,285% in as little as two days. Don’t let this become another “what-if” moment. Watch his free briefing now.
Forget the Sci-Fi Version of Nanotech
When you hear the word “nanotech,” you may picture tiny machines assembling new materials atom by atom, like molecular Lego.
But that’s the science fiction version. In real life, nanotech is the engineering that makes it possible to:
- Vaporize tin into 220,000-degree plasma to print circuits a few atoms wide
- Deposit more than 100 films onto a chip, each one measured in tenths of a nanometer
- Build sensors that detect flaws the width of 10 atoms
Nowhere is nanotech more critical right now than in AI.
Every AI model – every chatbot, every image generator, every system that drives an autonomous car – runs on chips. And every new generation of AI demands more powerful chips than the last.
But you can’t just make chips bigger. Physics won’t allow it. More surface area means more heat – and at the speeds AI demands, even a few extra millimeters of distance between transistors creates lag that slows everything down.
The only way to make chips more powerful is to pack more transistors into the same space.
Today’s most advanced chips hold more than 100 billion transistors. Each one is just a few nanometers wide.
For perspective, a nanometer is to a marble what a marble is to the Earth. Getting to that scale – and manufacturing reliably at it, billions of times over – requires an entire ecosystem of highly specialized companies.
Here are three that need to be on your radar.
1. ASML (ASML)
The machines that print those impossibly small patterns are built by a Dutch company called ASML (ASML).
Each one weighs as much as two Boeing 737s, ships in 250 crates, and takes 250 engineers six months to assemble.
Inside, a high-powered laser fires 50,000 times a second at tiny droplets of molten tin. The first pulse flattens each droplet. The second vaporizes it at 220,000°C – roughly 40 times hotter than the surface of the Sun.
The resulting plasma emits a burst of extreme ultraviolet light that etches circuit patterns onto silicon.
The light is guided by mirrors polished so precisely that if you scaled one to the size of Germany, the tallest bump on its surface would be 0.1 millimeters.
There are only a few hundred of these machines on Earth. Each costs north of $300 million. And every advanced AI chip in existence was made on one.
2. Applied Materials (AMAT)
Once a pattern is printed, chipmakers have to build the next pattern on top of it.
Some layers conduct electricity – they’re the wiring that connects billions of transistors to each other. Other layers insulate. At atomic scales, electrical signals don’t stay neatly in their lanes. Without barriers, signals cross, and the transistor dies.
Applied Materials builds the machines that deposit these layers inside a vacuum chamber, one atomic film at a time. Gases are broken apart and reassembled on the wafer’s surface, building up films just a few atoms thick.
A single chip can require more than 100 of these layers, each one deposited with a precision measured in angstroms – tenths of a nanometer.
3. KLA Corporation (KLAC)
After a wafer is printed, it has to be checked for flaws.
The problem is the defects that matter at these scales are invisible to any optical microscope.
So chipmakers use electron beams instead. KLA’s systems fire a stream of electrons at a finished wafer and measure the scatter. Using this method, they can detect defects as small as 1 to 3 nanometers – the width of about 10 atoms.
Without nanotech quality-control tools, the chips that power ChatGPT and autonomous vehicles would roll off the line with hidden flaws no human eye could catch.
How to Play It
The three stocks listed above are a great place to start your research. But I recommend putting AMAT at the top of your shopping list… especially if you’re looking for a short-term trade.
Each of these stocks has a bullish forecast according to our Predictive Alpha AI-trading model. But AMAT’s forecast is the most bullish of the three.
By March 12, our model projects a gain of 9.3%. And the historical target accuracy for this model is above 78%, a relatively high reliability based on past performance.
AMAT is also in a Green Zone on our Short-Term and Long-Term Health indicators, another sign it’s in a bullish uptrend.
These bullish indicators don’t guarantee a rising price for AMAT. But together with its foundational role in powering the AI boom, they further stack the odds of success in your favor.
Keith Kaplan
CEO, TradeSmith
P.S. Don’t worry if you’re not yet a TradeSmith Platinum member. I cover a lot of the themes our new Global Trend Tracker tool surfaces using real-time market data over at my X account.
Yesterday, I posted about how you can unlock passive income streams from another AI-related play – natural gas. AI datacenters require massive electricity loads. But production alone isn’t enough – you also need transportation.
That’s where natural gas pipelines come in. And I found an ETF that covers pipeline operators that pays out a 7.5% annual yield. Check it out here.
