What Is Cloud Computing (And Why Does It Matter)?

What the cloud physically is

IaaS, PaaS, and SaaS

Why companies left server closets

note

Why on-premises servers became a burden

Running servers in-house means paying for:

• hardware purchases
• electricity and cooling
• physical security
• backups and disaster recovery
• software updates and patching
• spare machines for failures
• staff on call

Why cloud won

Cloud providers spread costs across millions of customers. That creates:

• lower upfront cost
• faster setup
• easier scaling
• better geographic reach
• built-in redundancy options

Real-world example

Netflix announced its cloud migration in 2010 and later became one of the best-known large-scale AWS users. The point was not just cost. It was resilience and speed of change.

illustration

diagram

note

Elastic scaling

Elastic scaling means resources grow or shrink with demand.

If a shopping app gets 10 times more traffic on Black Friday, the cloud can add more servers. When traffic drops, those servers can be removed.

That flexibility is one of the biggest reasons cloud computing matters.

Serverless and the moving edge

Choosing the right cloud model

Transcript

Welcome to Slate. Today we're looking at What Is Cloud Computing (And Why Does It Matter)?. We'll cover What 'the cloud' actually means physically, IaaS, PaaS, SaaS — the three layers of cloud services, Why companies moved from servers in closets to AWS, and Serverless, edge computing, and where cloud is heading. Let's get into it.

The cloud is not a place in the sky. It is someone else’s computers, running in real buildings. Here’s the physical picture: racks of servers, storage arrays, network switches, backup power, and cooling systems packed into data centers. Amazon Web Services launched in 2006. Microsoft Azure followed in 2010. Google Cloud Platform arrived in 2008, then expanded under that name in 2011. These companies rent pieces of that giant machine over the internet. Think of it like a city utility grid. You do not build your own power plant to turn on a lamp. You plug into the grid and pay for what you use. Cloud computing works the same way for compute, storage, and networking. A laptop in a cafe can connect to a server in Northern Virginia, Dublin, or Singapore in milliseconds to hundreds of milliseconds, depending on distance and routing. The key shift is ownership. With on-premises computing, a company buys servers, installs them in a closet or data room, and becomes responsible for everything: hardware failures, patches, power, cooling, and spare parts. In the cloud, the provider owns the building and the machines. The customer rents capacity and focuses on the software.

Cloud services come in layers. The easiest way to see them is to ask one question: who manages what? Infrastructure as a Service gives you the raw building blocks. Platform as a Service gives you a managed runtime. Software as a Service gives you a finished application. With Infrastructure as a Service, or IaaS, you rent virtual machines, disks, and networks. Amazon Elastic Compute Cloud, launched in 2006, is the classic example. You choose the operating system, install software, and patch it yourself. It is like renting an empty storefront. With Platform as a Service, or PaaS, the provider manages the operating system, runtime, and often scaling. Heroku is a well-known example. You push code, and the platform runs it. That is like renting a furnished kitchen. You still cook, but the stove, plumbing, and electricity are already there. With Software as a Service, or SaaS, you just use the application. Gmail, Salesforce, and Microsoft 365 are SaaS. You do not manage servers or deployment. You open the app and work. These layers are not just labels. They define responsibility, speed, and control. More control usually means more work. More managed service usually means less maintenance.

Before cloud computing, many companies ran their own servers in office closets or small server rooms. That worked until demand grew. Then the hidden costs showed up. A server is not just a box. It needs power, cooling, backups, security, replacement parts, monitoring, and someone awake when it fails at 2 a.m. A famous example is Netflix. It moved major parts of its infrastructure to Amazon Web Services after a 2008 database corruption incident. Netflix wanted to scale globally and survive failures better than a single data center allowed. Cloud made that easier because capacity could grow and shrink with demand. The economics matter. If you buy hardware for peak traffic, much of it sits idle on ordinary days. That is like owning 20 delivery vans because of one holiday rush. Cloud lets you rent 5 vans most days and 20 only when needed. That is why startups loved it. They could launch without spending millions on hardware. Cloud also changed reliability. Providers run multiple availability zones, which are separate facilities in a region. If one building has trouble, traffic can move to another. That is much harder to do with one closet in one office.

Serverless does not mean there are no servers. It means you do not manage them. The provider runs the machines, starts your code when needed, and shuts it down when idle. AWS Lambda launched in 2014 and made this model famous. You pay for the time your code actually runs, often measured in milliseconds. That changes how teams build software. Instead of keeping a server waiting all day, you write small functions that react to events: an image upload, a payment, a database change. It is like a light switch connected to motion sensors. Nothing happens until there is a trigger. Edge computing pushes some of that work closer to users. Instead of sending every request to one faraway region, cloud providers place compute near the edge of the network. That can reduce latency for things like video delivery, gaming, or fraud detection. Cloudflare and AWS both offer edge services for this. The future is not one model replacing all others. It is a mix. Large databases may stay in central regions. Fast user interactions may move to the edge. Event-driven logic may run serverless. Cloud is becoming a toolkit, not a single place.

Cloud is not one decision. It is a set of choices. A startup building a new web app might use SaaS for email, PaaS for deployment, and a managed database in the cloud. A bank may keep sensitive systems on private infrastructure while using public cloud for analytics. A video app may put content delivery at the edge and core logic in a central region. Here is the practical test. If your team wants maximum control and has operations expertise, IaaS may fit. If your team wants to ship quickly with less maintenance, PaaS or SaaS may be better. If a task is event-driven and short-lived, serverless can save money and simplify scaling. If users are far away from your main region, edge services can improve response time. The cloud matters because it changed the default. In the old model, every company had to become a small data center operator. In the cloud model, infrastructure is a service you can rent, resize, and discard. That lets teams focus on the product instead of the plumbing. The plumbing still exists. It is just farther away, more standardized, and shared at enormous scale.