Lesson summary
Cloud computing means computing services delivered over the internet on demand. For this lesson, keep the idea simple: the organization requests capacity as a service, uses it while it is needed, and can scale or release it without buying and installing new datacenter hardware first.
The exam is not asking whether a server is far away. A company-owned server that employees reach over a VPN can still be traditional infrastructure. The cloud cue is the service model: internet-delivered capacity, faster adjustment to demand, and usage-aligned spending.
The exam move
When a scenario asks you to define cloud computing, look for three signals before you chase product names:
- The computing service is delivered over the internet.
- Capacity can be requested, increased, decreased, or released when demand changes.
- Cost follows use more closely than owned hardware that sits idle after purchase.
If the answer choice is mainly about who maintains the operating system, whether the deployment is public or private, or which Azure compute product to choose, it is probably testing a neighboring objective.
When a stem says a service can scale up or scale out, keep the cue conceptual: scaling up changes the capability of a resource, and scaling out changes the number of resources. For this definition lesson, either cue points to demand adjustment, not uptime design.
Scenario cue table
| Scenario cue | What it usually means | Keep out of the answer |
|---|---|---|
| A team needs extra capacity for a launch week without buying servers | Cloud computing demand cue | Specific VM sizes, scale-set configuration, or availability zones |
| Employees connect remotely to a company-owned server | Remote access, not automatically cloud computing | Assuming every remote system is cloud |
| Resources are rented for a project and released when finished | Consumption-based cloud spending | Claiming costs vanish while resources remain allocated or consumed |
| A service scales up or out as demand rises | Scalability cue inside cloud benefits | High-availability wording unless the question asks about uptime |
| An answer choice focuses on public, private, or hybrid | Deployment model boundary | Turning the definition question into a cloud-model selector |
Decision flow
Use this flow to keep the answer at the right level. It starts with delivery, then tests demand and cost cues, then rejects adjacent detail traps.
Worked scenario
A retailer expects traffic to spike during a weekend sale. In a traditional datacenter plan, the retailer would need to buy enough hardware ahead of time, install it, and keep paying for that capacity even after the sale. In a cloud scenario, the retailer uses internet-delivered computing capacity for the event, scales to handle demand, and scales back or releases resources afterward.
The correct exam reasoning is not "Azure Virtual Machines are available" or "the workload is highly available." The important cue is that computing capacity is delivered as a service and aligned to demand. If a distractor says the company simply connects to its own server from another location, that is remote access by itself, not the full cloud computing model.
Keep adjacent details out
This concept gives you the definition boundary. Shared responsibility, public/private/hybrid cloud, serverless, IaaS/PaaS/SaaS, and Azure compute services all matter later, but they are not the answer to a basic "define cloud computing" item. Treat those details as follow-up questions unless the stem explicitly asks for them.
Quick recap
Cloud computing is internet-delivered computing capacity used on demand. On the exam, prefer answers that connect internet delivery, elastic capacity, and usage-aligned spending. Reject answers that only describe remote access, generic virtualization, or an Azure product list.
Use the targeted assessment after this lesson to check whether you can spot the definition cue before adjacent Azure details pull you off scope.