What Does “X-Nanometer Chip” Actually Mean? | Jadewell Family Office

Recently, a client asked us:

“Hey, I heard TSMC’s 2‑nanometer process started mass production late last year. The news keeps talking about 2nm and 3nm every day… but what do these numbers actually mean?”

So today, let’s explain it in the simplest way possible. 🚀

_______________________

▋ Remember ET? 👽👉👈🧑

In the classic movie E.T., the alien’s glowing finger touches a human finger.

Now, imagine those two fingers as transistors.

On a tiny chip, countless transistors communicate using 0s and 1s:

🟢 Current flows = 1

🔴 Current is blocked = 0

So in your imagination, when ET’s finger connects with the human finger, that’s a 1.If they can’t connect, that’s a 0.

How do we control whether the two fingers can touch?

Very simple — put a gate between them.

_______________________

▋Early Chip Manufacturing: The “Rolling Shutter” Era 🏗️

In early chip designs, the gate worked like a rolling metal shutter, dropping down from above to block the connection.

The narrower the gate, the faster the “fingers” could connect, and the more efficient the chip became.

That’s why chip processes shrank from:

0.18 micrometers → 90 nanometers (nm) → 65nm → 45nm

Here, “micrometers” and “nanometers” refer to the width of the gate.

(For reference:

1 micrometer = 0.0001 cm

1 nanometer = 0.0000001 cm)

_______________________

▋When That Wasn’t Enough, We Switched to “Fins”🦈✨

After 40 years of using the rolling‑shutter design, problems appeared below 20nm:

• The gate became too narrow

• It couldn’t fully close

• Leakage became severe ⚡😵

  
  

Scientists came up with a clever idea: increase the contact area. 🔎

What does that mean?

Originally, ET and the human only needed their fingertips to touch.But scientists changed the rule — now they must touch their entire palms to conduct electricity. 👽✋🧑

With this larger contact area, the flat “rolling shutter” could be upgraded into a 3‑sided gate:

• top

• left

• right

This 3D structure looks like a shark fin drifting in the ocean, so it’s called:

FinFET — the fin‑type transistor.

Thanks to this breakthrough, humans advanced from 20nm all the way to 3nm.

But here’s the important part:

Once FinFET arrived, the “X‑nanometer” naming stopped reflecting any real physical dimension.

It became a marketing label — companies can call it whatever they want.

So is TSMC’s 3nm equal to Samsung’s 3nm?

Not really. They’re no longer directly comparable.

_______________________

▋But Now Even “Fins” Aren’t Enough Either…😮‍💨

FinFET covers three sides, but the bottom is still wide open.

So in the most advanced 2nm processes, a brand‑new technology appears:

✨ GAA — Gate‑All‑Around ✨

In simple terms, it wraps ET’s finger from all directions — top, bottom, left, right — forming stacked nanosheets.

Advantages:

• Lower leakage

• Higher density

• Perfect for AI’s massive logic workloads

This approach is extremely challenging, but it will be the core technology of chipmaking for the next decade. 🤖⚙️

_______________________

▋ Sounds Familiar? Isn’t This Like HBM Memory?

Exactly! Sharp readers may have noticed.

High Bandwidth Memory (HBM), widely used in AI systems, already uses layer‑by‑layer stacking.

So why can’t logic chips be stacked the same way?

Because:

Memory chips:

• Low power

• Low current

• Regular structure

Logic chips:

• High power

• High current

• Huge heat

• Complex structure

_______________________

▋Investing Starts With Fundamentals 📈💼

Everyone says you should study fundamentals before investing —but who has that much time?

Why not leave it to Jadewell Family Office?

We explain the markets and technology in the simplest, most engaging, and most digestible way.

Helping you understand technology, understand the markets, and understand the future. ✨

Risk disclosures and disclaimers: https://www.jadewellfo.com/disclaimer