Clicked Gallery

What is a cache?

Highlighted from a real engineering doc. Explained by Clicked.

Used in a sentence

Engineering Notes · AI Systems

The app felt sluggish on first launch, but once the cache warmed up, every screen loaded instantly.

The reader highlighted one word in the docs. Clicked explained the technical term “cache” in simple terms:

Explained in three depths

Same facts, different vibe — Slang mode 😎

The Clicked way

●○○

Overview

A cache is a small store of copies kept close to where they will be needed. When the same thing is requested again, the copy is served on the spot instead of being fetched from the slower original source. Your browser keeps copies of pages you visit, and websites keep copies of answers they just computed. The copies are disposable: clearing a cache loses nothing permanent. It just makes the next request slow again.
●○○

Overview

A cache is your device deciding it's not making the same slow trip twice. It kept a copy from last time, so round two loads instantly and everyone assumes the internet got faster. It didn't. The copy was just already here. And since it's only a copy, deleting it breaks nothing: you're not losing the thing, you're losing the shortcut. 😎

A quick take — often all you need.

●●○

Detail

A cache is a store of copies placed close to where the work happens, and it exists because reaching the original source is slow. A request goes to the cache before it goes to the source. If the copy is there, a cache hit, it is served immediately. If not, a cache miss, the request goes the slow way to the source and a copy is saved for next time. This pays off because requests repeat: the page you reload, the file a program opens over and over. Two limits follow from the design. A cache is small, because fast storage costs more than slow storage and there is less of it. When it fills up something has to go, usually whatever was used least recently. And a copy can go stale. The source moves on, the copy does not, and the cache keeps serving yesterday's version, because it cannot tell that anything changed. Every copy therefore carries an expiry date, and once it passes, the cache fetches a fresh one. Emptying the cache does the same thing immediately, which is why tech support suggests it for a site showing outdated content. The same idea runs at every scale. A processor caches data it just used and your phone caches app content.
●●○

Detail

A cache is your device keeping a copy of whatever it just fetched, so next time it serves the copy instead of asking again. That one move explains why everything is fast the second time, and why "clear your cache" is the most prescribed medicine in tech support. Both, same move. The awkward part shows up when the original changes, because nobody tells the copy. There's no group chat where a website announces its update, so your browser cheerfully serves last week's version while the real page has moved on. Nothing is broken and nothing is lying. The copy is just old, and your browser has no way to know that. Copies do expire on their own, which is the polite fix, but the expiry date was set by whoever built the site, and they may have been generous. Deleting the copies skips the wait: nothing is left to serve, so everything gets fetched fresh, and the mysterious bug goes with it. The price is the same every time. First visit to every site is slow again, because you just threw away every shortcut you had. 😎

Want more? One click digs deeper.

●●●

Analogy

A cache is a pantry, and the supermarket is the original source. Nobody drives to the store every time a recipe needs flour: the flour came home once, and the pantry keeps it close to the stove. A pantry is small, so it carries what gets used, not everything the store sells, and stocking one new thing eventually pushes out another. Its contents can also outlive their usefulness. Something sits past its date until someone checks. And nothing in it is the only one of its kind. Empty every shelf and you lose no food, only the convenience, because the store still sells all of it.
Deepest level reached.
●●●

Analogy

A cache is your bedside water glass. The tap has all the water, but the tap is a whole journey at 2 a.m., so past-you left a glass within arm's length. That's the entire trick: the thing you'll want next, positioned where you'll want it. The glass is tiny compared to the tap, and it should be, because its job is nearness, not supply. Leave it too long and it goes flat and dusty, at which point you don't repair anything. You tip it out and refill it. Nothing of value was lost. The tap still has every drop. 😎
Deepest level reached.

Unfamiliar concept? A real-world example makes it click — fresh analogies on tap.

AI explanations may contain errors · Not professional advice

Formal definition — The same term, explained the usual way

A cache is a hardware or software component that transparently stores copies of data so that subsequent requests for that data are served faster than by recomputing or refetching from the underlying source. Requests satisfied from the cache are hits; the remainder are misses and are forwarded to the backing store, typically populating the cache on return. Because capacity is limited, entries are evicted under a replacement policy such as least recently used, and consistency with the source is managed through invalidation and expiry (TTL). Caching appears at many levels, including CPU caches, operating system page caches, browser caches, and content delivery networks.

Want Clicked to explain terms like “cache” directly in your browser — including on PDFs?

Add to Chrome — Free

50 free Explanations · No credit card required