Cloud computing
Renting computers over the internet instead of buying your own.
Cloud computing means using someone else's computers over the internet, paid by the hour or by use, instead of buying and running your own. The "cloud" is not in the sky: it is huge buildings full of servers (a data center) owned by companies like AWS, Microsoft and Google, which rent out slices of them.
What you rent ranges from a bare virtual machine (Amazon EC2), to file storage (Amazon S3), to a managed database, to a function that runs for a few milliseconds (AWS Lambda). You click a button or call an API, and the machine exists a minute later; delete it and the bill stops.
Why it matters: a startup can begin with one small server and grow to thousands without ever buying hardware. Netflix, your bank's app and most AI chatbots run in the cloud. The trade-off is that you depend on the provider, and the bill grows with use, so a mistake left running over a weekend can get expensive.
Keep in mind
- The three biggest providers are AWS, Microsoft Azure and Google Cloud.
- You pay for what you use: per hour, per gigabyte or per request.
- "Public cloud" means shared data centers; "on-premises" means servers you own.
- Elasticity, adding or removing capacity in minutes, is the whole point.
