The Dumbest Smart Thing in Your House
There's a tiny brain inside your phone, your laptop, your car — even the washing machine. It's called a processor, and here's the surprise: it can only do really simple things. It's just incredibly, ridiculously fast.
Your phone can find the way to a friend’s house, play a game with you, and video-call someone on the other side of the world. So it must have a pretty clever brain inside it. Right?
Here’s the surprise this whole lesson is about: the brain inside your phone isn’t clever at all. It can only do tiny, simple things — add two numbers, check which of two numbers is bigger, move a number from one place to another. That’s about it.
So how does it do all that amazing stuff? It isn’t smart. It’s fast. Ridiculously, unbelievably fast.
That brain is called a processor — a small chip that follows instructions, one tiny step at a time. And it might just be the dumbest smart thing in your house.
They’re hiding everywhere
Before we look inside one, let’s go hunting. Processors are tucked away in far more places than you’d think:
- your phone, laptop and tablet
- the games console and the TV
- the car — and not just one: a modern car has dozens of small processors, each looking after its own job, like the engine, the brakes or the windows
- the microwave and the washing machine, which have tiny, simple processors that follow their own short list of instructions: spin, stop, drain, beep. They’re nothing like a laptop, but they’re processors all the same.
- even the little gold square on a bank card. It’s the outside of a chip, and inside that chip is a tiny processor.
Here’s another surprise: even a single gadget usually has more than one brain. A laptop has its main processor, plus small ones quietly looking after things like the battery, the wifi and the storage. And the main processor is itself split into several parts called cores, and each core can follow its own list of instructions.
What’s inside: switches
If you could shrink down to the size of a speck of dust and walk across a processor, you’d find it covered in switches. Billions of them.
So a processor isn’t one switch. It’s a chip made of billions of transistors, all wired together.
And here’s the strange part. One switch knows nothing. It can’t add. It can’t count. It has no idea what a picture or a game is. It’s on, or it’s off, and that’s all it can ever do.
But wire a handful of switches together in just the right pattern, and they can add two small numbers. Wire together more, and they can compare numbers, or hold on to a number for later. Keep going — a new phone’s processor can have more than ten billion transistors — and you get something that can run everything a computer does. Not because any part of it is clever, but because there are so many parts, all doing their one tiny thing.
Dumb, but fast
Here’s roughly what every single step a processor takes looks like:
- add two numbers
- compare two numbers (which is bigger? are they the same?)
- move a number from one place to another
That’s the whole toolbox. Every app, every game, every video call is built out of enormous piles of steps like these.
So where does the magic come from? Speed. A processor takes billions of these steps every second.
If you’ve read How Big Is a Billion, Really?, you already know how huge that is. Imagine doing one processor step every second, never stopping — no meals, no sleep. A billion steps would take you more than 30 years. A processor does that many in about a second.
Be the processor
Time to feel it for yourself. First, you are the processor: follow the instructions one at a time and colour in a picture, square by square. Watch the step counter and the clock. After ten steps you can skip ahead — you’ll have got the idea.
Then hand the same job to two kinds of chip, and see what changes.
- 1. You (now)
- 2. One fast worker
- 3. Thousands of helpers
Mode 1 of 3: You are the processor
A processor only follows tiny instructions, one at a time. So be one. Do each instruction by tapping that square — or use the arrow keys and press Enter. What picture are you making?
The instructions
- Now: Square 3 → blue
- Later: Square 4 → blue
- Later: Square 5 → blue
Steps done
0 of 44
Time taken
0:00
Notice what didn’t change across all three. Every step was the same dumb job: colour this square. Nobody got smarter. The only things that changed were how fast the work was done, and how many workers were doing it at once.
Two kinds of worker
That’s exactly the difference between the two main kinds of processor.
A normal processor is called a CPU, short for central processing unit. Think of it as a kitchen with a few super-fast chefs. Each chef can cook almost anything: soup, then a cake, then a salad, one job after another. That makes a CPU great at doing lots of different jobs in order — which is most of what your phone or laptop does all day.
A graphics chip is called a GPU, short for graphics processing unit. It’s a kitchen with thousands of helpers. Each helper can only do one simple thing, like butter a slice of toast. But if you need a thousand slices buttered, a thousand helpers can each do one slice at the same time, and the whole lot is done in one go. That’s what you saw in the third part of the machine above.
Pictures need exactly that. A screen is made of millions of tiny squares called pixels, and each one needs the same small sum worked out to get its colour — many times every second. So screens are drawn mostly by GPUs.
And AI needs it too. Underneath, AI is gigantic piles of the same kind of simple sum. So AI runs mostly on those “thousands of helpers” chips, packed by the thousand into giant buildings full of computers called data centres. There are even newer chips built specially for AI, which take the helper idea further still.
Neither kind of chip is smarter than the other. The chefs are the right choice when there are lots of different jobs to do in order. The helpers are the right choice when there’s one simple job to do millions of times. Your phone and your laptop have both.
Coming up: how do you build one?
That leaves a puzzle. A processor has billions of switches on a chip about the size of your fingernail. How on earth do you make switches that small? And what are they made of?
Here’s one clue: mostly silicon, an element — one of the kinds of atom you can meet in Atoms: The Building Blocks. How you turn it into billions of switches is the next part of the story.
And the companies that design, build and sell all of this are some of the biggest in the world. That’s what How Big Tech Works is about.
Key words
Processor vocabulary
Card 1 / 6Front
Check yourself
Processors — quick check
Question 1 of 5What can a single transistor do on its own?
So the next time something on a screen seems clever, remember what’s really going on underneath: a chip full of switches, taking billions of tiny, dumb steps every second. Not smart. Just fast.