How processors work
The Billion-Switch Metronome Inside Every Device You Own
Your phone's processor makes more decisions every second than all the neurons in your brain fire in a lifetime.
The Idea
At the heart of every processor is a deceptively simple idea: electricity can be made to represent two states — on or off, one or zero — and if you combine enough of those states in the right sequence, you can represent anything: numbers, words, images, music, decisions. The magical part isn't the binary logic itself but the speed and miniaturisation at which it operates. The fundamental component is the transistor — essentially a microscopic switch that can be flipped by an electrical signal rather than a human hand. Modern processors contain tens of billions of these switches, each smaller than a virus, etched into silicon wafers through a photographic process of almost absurd precision. They are grouped into logic gates, which perform basic operations (AND, OR, NOT), and those gates combine into circuits that can add numbers, compare values, and route data. What synchronises all this activity is the clock — a crystal that vibrates at a fixed frequency, ticking billions of times per second. Every tick is an opportunity for transistors to change state. Clock speed, measured in gigahertz, is one reason a newer chip feels faster, but it's not the whole story. Modern processors are also deeply parallel: they contain multiple cores that work simultaneously, and they speculate ahead, predicting which instructions will be needed next and pre-computing them before they're asked. The processor isn't just fast — it's also, in a modest but real sense, anticipatory.
In the World
In 1965, Gordon Moore — then at Fairchild Semiconductor, later a co-founder of Intel — published a short paper observing that the number of transistors engineers could fit on a chip had roughly doubled every year. He predicted this would continue. It did, more or less, for over half a century, and the observation became known as Moore's Law. The implications were staggering. The ENIAC computer, completed in 1945, contained around 18,000 vacuum tubes (the transistor's bulkier predecessor), weighed 30 tonnes, and filled an entire room at the University of Pennsylvania. It could perform roughly 5,000 additions per second. The processor in a current high-end smartphone contains more than 16 billion transistors, weighs a few grams, and executes hundreds of billions of operations per second. But Moore's Law is now straining against physics. Transistors on leading-edge chips are just a few atoms wide — at that scale, electrons start tunnelling through barriers they should, classically speaking, be blocked by. Leakage and heat become serious problems. The industry's response has been to stack chips vertically, design specialised processors for specific tasks (like the dedicated AI chips in modern phones), and explore materials beyond silicon. The straightforward doubling has slowed, but the ingenuity trying to sustain progress in its absence might be even more impressive.
Why It Matters
Understanding even the broad strokes of how processors work quietly changes how you think about the devices you live with. A laptop that feels 'slow' isn't failing — it's performing billions of perfectly logical operations; the bottleneck is usually memory access, not raw computation. When you hear that AI requires enormous computing power, you can picture what that actually means: not cleverness, but an almost incomprehensible number of multiplication operations, executed in parallel, across racks of specialised chips. There's also something genuinely humbling about the scale of it. The transistors enabling you to read this are so small they require electron microscopes to see, yet their collective behaviour produces recognisable text, music, conversation. It's emergence in the most literal sense — nothing in a single transistor looks anything like intelligence or creativity, yet aggregated across billions, at billions of cycles per second, the result convincingly mimics both. The next time a device surprises or frustrates you, the processor underneath is not magic and not a mystery. It's logic, repetition, and coordination — operating at a scale that strains human intuition but follows rules you could, in principle, trace all the way back to a switch flipping on and off.
A Question to Ponder
If the gap between a transistor and a thought is purely a matter of scale and organisation, what does that imply about the gap between a processor and a mind?
Get a new one of these every morning.
Start learning with Thinkable