Nine short lessons, from “what even is a server?” to your own setup. Click through at your own pace. Nothing to install, nothing to break.
Lesson 1
What even is a server?
A server is just… a computer. That’s it. The only difference from the laptop on your desk is that it lives in a big, secure, air-conditioned building (a data centre), it never gets switched off, and its whole job is to serve things up to other people over the internet.
When someone types yoursite.com, their computer asks your server for the page, and the server sends it back. It does that for everyone, all day, every day.
The office analogy (we’ll use it all the way through): think of a server as a small office that’s always open. People phone up asking for things; the office does the work and posts back the answer. Everything else in this course is just what’s inside that office.
Takeaway: a server is an ordinary computer that never sleeps and works for lots of people at once.
Lesson 2
Inside the box — click each part
A server is built from a handful of parts, and each one maps neatly onto our office. Tap a part below to find out what it does.
Pick a part ↑
Each part of the server does one job in our office. Tap any of them to see which.
Tip: the brass box (GPU) is the thing we’re asking the server people to add.
Takeaway: CPU = worker, RAM = the desk, Storage = the warehouse, GPU = a hall of extra workers for AI, Network card = the front door.
Lesson 3
CPU vs GPU — why AI needs the second one
These are the two “thinking” chips, and they think in completely different ways.
The CPU is one very clever, very fast worker. It does jobs one after another, brilliantly. It runs your whole website perfectly well.
The GPU is a hall of hundreds of simpler workers, all doing the same kind of sum at the same instant.
Here’s the key fact: AI “thinking” is thousands of tiny sums happening together. Give that to one clever worker and they’re there all day. Give it to the hall and it’s done in a blink. That’s why AI needs a GPU — and why a server without one can’t run our assistant at any usable speed.
Picture it: ask one accountant to add up 10,000 receipts and go for lunch — they’ll still be going when you’re back. Hand one receipt each to 500 people and it’s done before you’ve sat down. For AI, we want the room of 500.
Takeaway: CPU = one genius doing jobs in turn. GPU = a crowd doing one job together. AI is a crowd job.
Lesson 4
RAM vs Storage — the desk and the warehouse
People mix these two up constantly, but they’re completely different, and the difference is just desk vs warehouse.
RAM (the desk) is small but lightning fast. It’s where the computer keeps whatever it’s working on right now. Clear the desk (switch off) and it’s wiped — that’s fine, because it was only today’s working papers.
Storage / disk (the warehouse) is huge but slower. It’s where everything is kept when it’s not in use — every document, every file, the database, the AI’s brain-file. Switch off and it’s all still there.
You want enough desk so work flows, and a big, growable warehouse so you never run out of shelf space. (The GPU, remember, has its own little desk too — that’s the “VRAM” and the “24GB” number.)
Takeaway: RAM = a small fast desk that empties when you switch off. Storage = a big warehouse that keeps everything forever.
Lesson 5
How a click becomes a web page
What actually happens the moment someone opens a website? Press Play and follow the journey.
👤
Someone’s browser
🌐
The internet
🏢
Your server
🧠
App does the work
🗃
Database fetches data
📧
Page posted back
Press Play to send a request from someone’s browser to a website and watch it travel.
Takeaway: a click travels out across the internet to the server, the server does the work and fetches data, then posts the finished page back — usually in a fraction of a second.
Lesson 6
The software layers (what’s running on the box)
The parts in Lesson 2 are the hardware. On top of them sits software, in neat layers — like floors of a building, each one standing on the one below.
The operating system (the ground floor) — runs the whole machine. On our server it’s a version of Linux.
The web server — the receptionist that answers incoming requests from the internet.
The engine (software like PHP) — the machinery that builds each page and runs the logic.
The database — the giant, well-organised filing cabinet that stores every company, filing and record.
Your app or website — the top floor: your actual product, sitting on all of the above.
When you update your app or website, you’re only changing that top floor. The floors beneath keep everything standing.
Takeaway: hardware is the building; software is the floors inside it. Your product is the top floor, resting on an operating system, a web server, an engine and a database.
Lesson 7
“The cloud” and renting servers
“The cloud” sounds mystical but means something dull and reassuring: someone else’s computers, in someone else’s building, that you rent by the month. You get the power without buying and babysitting the hardware yourself. There are three flavours worth knowing:
Shared hosting — you rent a desk in a big shared office. Cheap, but you’re sharing with strangers and can’t change much.
A VPS (Virtual Private Server) — you rent a whole private room with a lockable door. Your own space, your own rules. Plenty of business sites run on one of these.
Dedicated — you rent the whole building. Most power and privacy, most cost.
A GPU request is essentially: “add a hall of AI workers to my private room, and make sure it’s mine, not shared.”
Takeaway: the cloud = renting computers by the month. We’re on a VPS (a private room); we’re asking to add a dedicated GPU to it.
Lesson 8
Putting it together: your setup
Now the whole picture, in the words you’ve just learned:
A typical business site lives on a VPS — a private rented room in a data centre.
Inside it: a CPU running the site, some RAM (the desk), a disk (the warehouse) holding every file and record, and a database keeping it all organised.
On the top floor sits your own app or website — the bit your customers actually use.
If you want AI on it, what’s missing is the GPU — the hall of workers — so it can think fast and privately. That would be the upgrade.
And if you hold sensitive data, you keep the AI on your own kit — nothing goes to an outside service.
So next time someone talks about GPUs, VRAM, VPS, storage and backups — you’ll know each one is just a worker, a desk, a private room, a warehouse and a safety copy. You’re not out of your depth.
Takeaway: you now understand a real server. Adding AI is one new part (the GPU) plus room to grow — ideally kept in-house.
Check yourself
A quick five-question quiz
No pressure — instant feedback on each. See how much stuck.
0 / 5
You’re done! That’s genuinely most of what you need to hold your own in a server conversation. Keep the upgrade guide handy for the specifics.