The cloud is somebody’s computers
The word “cloud” suggests something vague and weightless, but the reality is ordinary and physical. Cloud computing means using computers, storage and software that are located elsewhere, in large buildings called data centers, and reached over the internet. Instead of buying and running your own machines, you use someone else’s, usually paying for what you use or a subscription.
The name comes from the habit, in network diagrams, of drawing the internet as a cloud shape: a part of the picture whose inner workings did not matter to the person drawing.
What is inside a data center
A data center is a facility full of racks of servers, which are computers designed to run continuously without a screen or keyboard. Around them are storage systems, networking equipment, power supplies, backup batteries or generators, and cooling systems to carry away heat. Large operators typically build facilities in several regions so that if one location has a problem, work can move to another.
The servers are often not used one machine per customer. Through a technique called virtualization, one physical server can be divided into several virtual machines that behave like separate computers, each isolated from the others. A related approach, containers, packages a program with what it needs to run in a lighter form. This sharing is what makes the cloud efficient: capacity can be allotted to customers who need it and released when they do not.
Who runs it
The cloud is run by companies that operate data centers and rent out access. Some are very large providers that offer a wide range of services to businesses and individuals. Others are smaller or specialized. Many organizations also build their own private clouds in their own data centers, and some mix both approaches, which is called a hybrid setup. Because these are commercial services, the terms, prices and locations of data differ from one provider to another, and they change over time.
Three common ways to rent
People often describe cloud services in three layers:
- Infrastructure as a service. You rent basic building blocks, such as virtual servers and storage, and manage the software yourself. It is like renting an empty workshop with power and tools.
- Platform as a service. The provider also manages the operating system and runtime, so developers can focus on their own application code.
- Software as a service. You simply use finished software through a browser or app, such as web-based email, document editing or photo storage. This is the form most ordinary people use.
These labels are a way of describing who is responsible for what; real products do not always fit tidily into one of them.
What changes for ordinary users
Many everyday activities rely on the cloud, even if you never see the word. Photos from your phone can be backed up automatically, and documents can be edited from several devices. Streaming services, online banking, maps, messaging apps and online games run on remote servers. Voice assistants and many features that use language models send your request to powerful remote computers because the device in your hand could not do the work alone.
The result is that:
- Your files can follow you across devices.
- A lost or broken phone need not mean lost photos, if they were backed up.
- Software updates and fixes happen on the provider’s side.
- You often need less powerful hardware, since heavy work runs elsewhere.
- Collaboration, such as several people editing one document, becomes much easier.
What changes for businesses
For organizations, cloud computing turns a large upfront purchase of equipment into an ongoing cost, and allows capacity to grow or shrink quickly. A small shop can use the same kind of services as a big company, which is part of why small business tools have become widespread. Large-scale data science work, which can need a lot of computing for a short time, is another case where renting is often more practical than owning.
Trade-offs worth knowing
The cloud is convenient, but there are real downsides.
- You need a connection. Without internet access, cloud-only features may not work. Many apps keep an offline copy for this reason.
- Outages happen. Even large providers have occasional failures, and when a widely used service goes down, many websites can be affected at once.
- Privacy and control. Your data sits on equipment you do not control, under the provider’s policies and the laws of the place where it is stored. Reading the settings and terms of services you use, and thinking about online privacy, is sensible.
- Security is shared. Providers protect their data centers, but your account is still a target. A strong, unique password, ideally with a password manager, and two-factor authentication make a real difference.
- Costs can creep up. Pay-as-you-go pricing is flexible, but it can add up if usage is not watched.
- Switching can be hard. Moving large amounts of data or software from one provider to another can be difficult or costly, which is sometimes called lock-in.
- Energy use. Data centers use substantial electricity and water for cooling. Large providers often argue that shared, efficient facilities use less than many small private rooms, but the net effect depends on the details.
Is “the cloud” the same as the internet?
No. The internet is the network that connects devices. The cloud is a set of services that run on computers attached to that network. You can use the internet without using the cloud, for instance by sending a message directly to a friend, though most modern services involve both.
Key points
The cloud means renting computing, storage and software from remote data centers rather than owning them. Providers divide big servers among many customers, and ordinary users meet the cloud as photo backup, web apps and streaming. The benefits are convenience and flexibility, and the costs are reliance on a connection, less direct control over data, and the need to look after account security.