What the cloud is
The cloud is a data centre you rent over the internet
Cloud computing delivers computing as a service: someone else owns the machines, and you rent what you need, when you need it.
Start with what the cloud physically is. Here is you, with a laptop. You send a request over the internet, and about a minute later a server is running and waiting for you. That server lives in a building owned by a provider such as Amazon, Microsoft or Google. The building holds the same four things any computer room holds: processors and memory, disks, a network, and software services built on top. The provider bought it all, and pays for the power, the cooling and the staff. You get a slice for as long as you want, and the bill counts exactly what you used. Nothing about the hardware is new. What is new is who owns it, and that asking for it takes a request, not a purchase.
What the cloud is · Going deeper · 1 of 6
From an experiment in 2006 to the default
Amazon rented out storage in March 2006 and servers that August. Google and Microsoft followed, and each later step, from platforms to functions to AI models, handed the provider one more job.
What the cloud is · Going deeper · 2 of 6
Five traits make a service cloud
The NIST definition: on-demand self-service, broad network access, resource pooling, rapid elasticity, measured service.
So what exactly counts as cloud? The standard answer has five parts. On-demand self-service: you start what you need yourself, with no ticket and no phone call. Broad network access: you reach it over ordinary networks, from a laptop, a phone or another server. Resource pooling: many customers share one large pool of machines, and you do not choose which machine you get. Rapid elasticity: what you have can grow and shrink quickly with demand. Measured service: your use is metered, shown to you, and billed. A service needs all five. A server rented from a hosting company on a yearly contract is someone else’s computer, but it is not cloud.
What the cloud is · Going deeper · 3 of 6
NIST’s whole model: 5 traits, 3 service models, 4 deployment models
The same two-page standard also names the service models (IaaS, PaaS, SaaS) and the deployment models (public, private, community, hybrid).
What the cloud is · Going deeper · 4 of 6
A rented server is not cloud
Traditional hosting fails self-service, elasticity, pooling and metering. The five traits are how you tell the two apart.
What the cloud is · Going deeper · 5 of 6
Public, private, hybrid, multi-cloud: where the machines sit
A deployment model says whose hardware it is and who shares it. Most organisations end up hybrid.
Cloud is not one place. Public cloud is the provider’s hardware, shared with their other customers. It is what most people mean by the cloud, and it is the default. Private cloud is the same way of working, self-service and elastic, on hardware used by one organisation. Hybrid joins the two with a private link, so a system in your building and a system at the provider work as one. Multi-cloud means using more than one public provider, sometimes by design and often because a company bought another company.
What the cloud is · Going deeper · 6 of 6
Why anything stays on premises, and how the cloud reaches it
“On premises” means in a building you control. Four reasons keep systems there, and each provider sells a way to extend its cloud to them.
“On premises” means in a building you control. Four reasons keep systems there, and each provider sells a way to extend its cloud to them. Reasons to keep a system where it is. Latency A machine on a factory floor or a trading system needs an answer in a few milliseconds, and the nearest cloud region is too far away. Law and contracts Some data must stay inside a country, a building or a certified facility (data residency). Hardware already paid for A data centre bought two years ago is a sunk cost. Moving before it wears out means paying twice. Licences and old systems Software licensed per physical processor, or a mainframe nobody dares to touch, can cost more to move than to keep. The provider comes to you. For exams: hybrid needs a connection between the two sides, either a VPN over the internet or a dedicated private line (AWS Direct Connect, Azure ExpressRoute, Google Cloud Interconnect). Lesson 6 covers both.
Reasons to keep a system where it is
- Latency A machine on a factory floor or a trading system needs an answer in a few milliseconds, and the nearest cloud region is too far away.
- Law and contracts Some data must stay inside a country, a building or a certified facility (data residency).
- Hardware already paid for A data centre bought two years ago is a sunk cost. Moving before it wears out means paying twice.
- Licences and old systems Software licensed per physical processor, or a mainframe nobody dares to touch, can cost more to move than to keep.
The provider comes to you
Owning servers
Owning means buying for the peak and paying for the idle
Owned capacity is fixed: sized for the busiest hour, and idle for most of the others.
To see why renting won, look at what owning costs. This line is one week of demand for an online shop. Most of the week it is low. On Friday there is a sale, and for a few hours demand is five times higher. If you own the servers, you have to buy enough for that Friday, plus a margin. That is the flat line on top. Everything between the two lines is hardware you paid for, powered and cooled, that is doing nothing. In this week, about a fifth of the capacity is used. That is not unusual. Measured across real company data centres, typical servers run at twelve to eighteen percent.
Owning servers · Going deeper · 1 of 2
Guess too low and the best day becomes the worst
Under-buying fails at the moment of highest demand, and physical servers cannot arrive in time to fix it. Fear of this is why owners over-buy.
Owning servers · Going deeper · 2 of 2
The server is only the first line of what a server costs
Total cost of ownership counts every line: space, power, cooling, network, licences, people and replacement. Compare that total with the cloud bill.
Elasticity
Renting follows demand: capacity rises and falls with the load
Elasticity is capacity that grows and shrinks automatically to match demand, so you stop paying for idle machines.
Now the same week, rented. The stepped line is the capacity you are paying for. It sits just above demand, hour by hour. When the Friday sale starts, a rule notices the load rising and adds servers. This is called scaling out. When the sale ends, the same rule removes them, and you stop paying for them that minute. This is scaling in. The result: most of what you pay for is now in use, where before it was about a fifth. Nobody forecast the sale, and nobody stayed up to watch it.
Elasticity · Going deeper
Scale up or scale out: a bigger machine or more machines
Vertical scaling swaps in a larger machine and has a ceiling. Horizontal scaling adds identical machines behind a load balancer and has none, which is why clouds favour it.
Capital and operating expense
A purchase becomes a bill: capital expense turns into operating expense
Owning is CapEx: pay first, use later. Renting is OpEx: pay as you go, for what you used.
The money changes shape too. When you own, you pay one large sum before the first customer arrives, and then running costs every month. The purchase is a guess, and you live with it for three to five years. Accountants call this capital expense. When you rent, there is no purchase. Each month you get a bill for that month’s use, so it rises in a busy month and falls in a quiet one. This is operating expense. The trade is real in both directions. Renting needs nothing upfront and you can stop any day, but the bill moves, so somebody has to watch it.
Capital and operating expense · Going deeper · 1 of 4
You pay for servers multiplied by time, in any shape
One server for 100 hours costs the same as 100 servers for one hour. Per-second billing makes finishing sooner free.
Capital and operating expense · Going deeper · 2 of 4
Pooling makes it cheaper: customers’ peaks do not line up
Resource pooling: one shared pool needs far less hardware than its customers would need separately.
Why can a provider rent you a server for less than it would cost you to own one? Take three customers. A shop is busiest at nine in the morning. A game is busiest in the afternoon. A reporting job runs at night. Each needs a hundred units at its own peak, so if each owns hardware, together they buy three hundred. Now put them on shared machines and add their demand hour by hour. Because the peaks fall at different times, the total never comes near three hundred. The pool needs well under half. The provider buys that much, keeps it busy around the clock, and shares the saving with you.
Capital and operating expense · Going deeper · 3 of 4
Economies of scale: every unit gets cheaper
A provider wastes less electricity, designs its own chips, buys in bulk and automates. Economies of scale turn that into lower prices.
Capital and operating expense · Going deeper · 4 of 4
Multi-tenancy: shared hardware, separate machines
A hypervisor slices one physical server into isolated virtual machines for different customers.
Agility
Minutes, not months: a new server is one request away
Agility: when resources arrive in a minute, experiments are cheap and so is undoing them.
Time changes as much as money. To get a new physical server, a company approves a budget, places an order, waits for delivery, then racks, cables and installs it. Ten weeks is typical. To rent one, you send one request and wait about a minute. On this scale that minute is far thinner than the line drawn for it. The gap changes how teams work. When a server costs ten weeks, every decision is a commitment. When it costs a minute, you can try ten ideas this afternoon, keep the one that worked, and delete the rest.
Agility · Going deeper · 1 of 3
Four ways in, one API behind them
The web console, the command line, an SDK and an infrastructure-as-code file all send the same signed request to the provider’s API.
Agility · Going deeper · 2 of 3
The six advantages on the AWS exam are these mechanisms, renamed
Each official advantage is one mechanism: variable cost, scale, elasticity, speed, no data centre, global reach.
Every provider sells these ideas under its own names, and exams test the names. Amazon lists six advantages. Trade fixed expense for variable expense is the purchase that became a bill. Benefit from massive economies of scale is pooling and bulk buying. Stop guessing capacity is elasticity. Increase speed and agility is the server that arrives in a minute. Stop spending money running data centres is the building you no longer own. Go global in minutes is the provider’s data centres around the world, which is lesson three. In an exam, find the mechanism in the story first. The phrase follows.
Agility · Going deeper · 3 of 3
The same ideas in Microsoft’s and Google’s words
AZ-900 speaks of the consumption-based model, high availability, scalability, reliability and predictability. Google’s exam speaks of agility, total cost of ownership and transformation.
When the cloud costs more
Renting wins when demand moves; a flat load can be cheaper to own
On-demand prices include a premium for the freedom to stop. Steady workloads should commit, or may be cheaper owned.
Cloud is not always the cheaper answer, and it helps to know where it stops being one. The rising line is renting on demand. Use nothing and you pay nothing. Use a machine every hour of the month and you pay the most. The flat line is owning, or committing to a machine for years. It costs the same however much you use it. The lines cross. To the left, where machines are busy only part of the time, renting wins easily. To the right, where a machine is busy all day every day, the flat line is cheaper. Most workloads live on the left. The ones on the right are why providers sell long-term commitments at a large discount.
When the cloud costs more · Going deeper · 1 of 3
Data in is free, data out is billed
Egress is the charge for data leaving the provider. It is small per gigabyte and large at scale.
When the cloud costs more · Going deeper · 2 of 3
Lock-in: the more a service does for you, the harder it is to leave
Portable building blocks mean more work for you. Provider-only services mean less work and a harder exit.
When the cloud costs more · Going deeper · 3 of 3
What else you give up, and what to do about it
Renting moves some risks to the provider and creates a few new ones. None is a reason to avoid the cloud; each is a thing to plan for.
Renting moves some risks to the provider and creates a few new ones. None is a reason to avoid the cloud; each is a thing to plan for. Control of the hardware You cannot walk in, choose the disk model or delay the provider’s maintenance. In exchange, you never replace a failed disk. Shared fate When a provider region has a bad day, thousands of companies do at once. On 20 October 2025 a DNS fault in a single AWS region disrupted services worldwide for hours. Spreading a system across zones and regions is lesson 10. A bill that can surprise Anyone with access can start a hundred servers, and a forgotten test keeps charging. Budgets and alerts are lesson 9. Shared responsibility The provider secures the building and the hardware. What you put on it, and who you let in, is still yours to secure. That split is the next lesson. Leaving happens, rarely. Moving work back from the cloud is called repatriation. It makes sense for the far right of the chart above: large, flat and well understood.
- Control of the hardware You cannot walk in, choose the disk model or delay the provider’s maintenance. In exchange, you never replace a failed disk.
- Shared fate When a provider region has a bad day, thousands of companies do at once. On 20 October 2025 a DNS fault in a single AWS region disrupted services worldwide for hours. Spreading a system across zones and regions is lesson 10.
- A bill that can surprise Anyone with access can start a hundred servers, and a forgotten test keeps charging. Budgets and alerts are lesson 9.
- Shared responsibility The provider secures the building and the hardware. What you put on it, and who you let in, is still yours to secure. That split is the next lesson.
Leaving happens, rarely
The providers
Three providers hold about 63% of the market
AWS, Microsoft Azure and Google Cloud lead, and the three sell the same building blocks under different names.
Finally, who you rent from. Amazon Web Services was first, in 2006, and is still the largest with about thirty percent of the market. Microsoft Azure has about twenty percent, and is strongest in companies that already run Microsoft software. Google Cloud has about thirteen, with a reputation in data and AI. Everyone else shares the rest. The three leaders compete service for service, so what you learn on one transfers to the others. This course teaches each idea once, then shows you its three names.
The providers · Going deeper
Same building blocks, three sets of names
A virtual machine is EC2 on AWS, Virtual Machines on Azure and Compute Engine on Google Cloud. The idea is the thing to learn.
Lesson complete
You finished “What the cloud is”
Nicely done.