Cloud Computing Fundamentals

What cloud computing means

AWS describes cloud computing as the on-demand delivery of IT resources over the internet with pay-as-you-go pricing. Instead of purchasing servers, storage, databases, and networking equipment months in advance, a team can request the capacity it needs when it needs it. The important change is therefore not simply that a server runs in somebody else's data center; infrastructure becomes a service that can be provisioned, changed, measured, and released through software.

That service model shortens the distance between an idea and usable infrastructure. A development team can create an environment in minutes, evaluate it, scale it if demand grows, and remove resources when they are no longer needed. The same principle extends far beyond virtual servers: AWS offers managed databases, storage, analytics, security, integration, and AI services that reduce how much underlying infrastructure a customer has to operate directly.

Cloud versus traditional capacity planning

In a traditional data-center model, capacity decisions often happen before actual demand is known. Hardware must be selected, purchased, delivered, installed, powered, secured, and maintained. Because lead times can be long, organizations frequently provision for a forecasted peak. If the forecast is too high, expensive capacity sits idle; if it is too low, users may experience shortages until more equipment arrives.

Cloud infrastructure changes that relationship. Capacity can be adjusted much closer to real demand, and many services support automation that reacts to load. This does not eliminate architecture or capacity planning: quotas, service characteristics, performance requirements, data location, and failure modes still matter. What changes is the speed and granularity with which the plan can be executed and revised.

• Traditional infrastructure emphasizes ownership and long procurement cycles; cloud infrastructure emphasizes service consumption and programmable provisioning.
• Cloud capacity can be added or removed without buying physical hardware for every change.
• Measured consumption makes unused resources visible as an operational and financial concern rather than a sunk hardware purchase.

The characteristics that matter in practice

For day-to-day architecture, several characteristics appear repeatedly: resources are available on demand, capacity can be elastic, usage is metered, infrastructure is accessible through APIs and automation, and customers can choose from services at different levels of abstraction. These characteristics reinforce one another. Automation makes fast provisioning repeatable; elasticity makes variable capacity useful; metering connects resource decisions to cost.

A cloud platform also enables global deployment without an organization building its own facilities around the world. AWS Regions and Availability Zones provide the geographic and fault-isolation foundation for this. Managed services can further move operational work to AWS, although the customer always keeps responsibilities such as protecting its data, identities, and configurations according to the shared responsibility model.

What the cloud does not do automatically

Using AWS does not automatically make an application secure, highly available, inexpensive, or scalable. Those outcomes depend on architecture and operations. A single instance in a single Availability Zone can still become a single point of failure; an oversized resource that runs continuously can still waste money; broad permissions can still expose data. Cloud services provide capabilities, but customers decide how to assemble and govern them.

This distinction is useful throughout the Knowledge Base: first understand what a service or cloud capability makes possible, then understand the design choices required to realize its benefit. The AWS Well-Architected Framework provides a structured way to examine those choices across operational excellence, security, reliability, performance efficiency, cost optimization, and sustainability.