An introduction to cloud computing infrastructure

Cloud computing infrastructure has become the modus operandi for an increasing number of companies, including enterprises like Netflix, Capital One, and Coca-Cola. Many of these companies are replacing their traditional data centers and enterprise deployments of software with cloud based services.

But what about this approach is so enticing? And what benefits does it offer? What is the difference between cloud computing infrastructure and cloud computing architecture? All great questions. This post will cover the following topics to help you better understand why cloud computing infrastructure is so popular.

  • What is cloud computing? 
  • Cloud computing infrastructure vs. cloud computing architecture
  • A brief history of cloud computing
  • How does cloud computing work?
  • Four components of cloud computing infrastructure 
  • Cloud computing services
  • Benefits of cloud infrastructure
  • Limitations of cloud infrastructure

What is cloud computing?

Cloud computing infrastructure can be defined as the hardware, virtualization, storage, and network components needed to deliver on-demand computing services. Cloud computing provides the flexibility to only pay for what you need (cue the Liberty Mutual jingle) at a given moment in time. Consumers and businesses can “rent” IT resources instead of owning and maintaining entire physical data centers.  

The idea of sharing resources dates back farther than you might think. The fundamental practices that led to cloud computing infrastructure date back over 70 years ago.

Cloud computing infrastructure vs. cloud computing architecture

Hardware, storage, networking, and virtualization are all key components of cloud computing infrastructure. It’s helpful to think of them as the materials needed for a construction project. A cloud computing architecture is the blueprint for how those materials are integrated to build cloud deployment models.

A brief history of cloud computing

The exact origin and common use of the term “cloud” is difficult to pinpoint, but by the 1990’s, it was being used to symbolize a network of undefined and infinite number of endpoints, or the empty space between a user and a provider.  

The origin of cloud computing concepts go as far back as the 1950s, when computers were so massive and expensive that companies had to create “time-sharing” schedules. Multiple users could access mainframes from remote workstations, called “dumb terminals,” with no processing power. 

In the 1960’s, MIT developed a primitive cloud using reels of magnetic tape for a military project called DARPA (the Defense Advanced Research Projects Agency). Just two to three people could access this early cloud environment simultaneously. This marks the beginning of virtualization, one of the key components of cloud computing infrastructure, although its definition has since been broadened. A few years laters, J.C.R Licklider helped shape another component of cloud computing infrastructure through his vision of a Galactic Network. 

How does cloud computing work?

And now, the moment you’ve been waiting for--cloud computing explained! Cloud computing is made practical by virtualization, which separates the physical hardware, network, and location for storage into virtual resource pools that can be accessed remotely. Rather than owning all of the infrastructure and the administration overhead, businesses can lease resources that meet their needs. This flexibility allows users to purchase the industry’s best cloud computing services (since they are constantly evolving), without the upfront or long term cost of owning and maintaining the infrastructure.   

Let’s take a look at the components of cloud computing infrastructure: hardware, storage, network, and virtualization.

Four components of cloud computing infrastructure


While corporations may be moving away from being “hardware companies”, each owning large data centers to meet their needs, the back-end equipment is still needed to actually “compute” the workloads. The owners of data centers have transitioned to cloud providers like Amazon, Microsoft, IBM and Google, who each own over 50 data centers (according to IEEE ComSoc report).  

Here is just a small list of some of the hardware that is needed in each of these geographical data centers to make cloud computing infrastructure possible:

  • Servers, processors, and memory
  • Network switches, routers, and cables
  • Firewalls and load balancers 
  • Cooling systems
  • Power supplies 


Cloud computing infrastructure provides data retention through three prevalent storage types: file, object and block. Each one is ideal for different use cases—let’s examine further. 

File storage allows for data to be collected into a single piece of information and presented in hierarchical user-friendly file folders. 

Object storage is used for large quantities of unstructured static data. It has three parts: the raw data as a distinct unit called an object, metadata used to give context to the object, and a globally unique identifier. Object storage in cloud computing is often used for media files and archived data that needs to scale but not be edited.  

Block storage is appropriate when low latency operations and high performance workloads are needed within cloud computing. Files are broken down into “blocks” of data and a unique identifier, but no metadata is stored. Data can be stored across multiple locations and environments and can be easily recovered if there is a hardware failure. 


Use case


File storage

File storage, local archiving, and data protection

Easy to organize files in an hierarchy

Block storage

Databases, mission critical apps, and RAID (redundant array of independent disks) volumes

Low latency and high I/O (input/output) performance


Web apps, big data sets, and backup archives

Cost effective extremely scalable 


A cloud computing infrastructure would not be complete without access to back-end resources. A network, composed of physical hardware like routers, switches, and high-speed cable are the foundation for which the virtual networks operate.  

Cloud computing infrastructure providers like Amazon, Google and Microsoft provide an abstraction from the physical network configuration by creating virtual local area networks (VLANs) and virtual private networks (VPNs). Administrators have the flexibility to define what traffic is sent over private networks (on-premise and private cloud), public cloud and multi-cloud deployments. 


Virtualization is a software technology that makes the computing services independent from the physical infrastructure through abstraction. While virtualization and cloud computing are often used interchangeably, they are quite different. The former is the underlying technology that makes resources “virtual”, while the latter is the delivery of the shared services so they are automated, scalable and self-serviced. 

Benefits of cloud infrastructure

Cost savings

Companies can reduce operating expenses by avoiding upfront costs that come with purchasing hardware and software. Simply put, cloud computing infrastructure allows the user to only pay for what they need.


Cloud providers have redundant availability zones designed for high availability, fault tolerance, and scalability; this scale is far greater than what most companies can implement themselves.  

Disaster recovery

It might not be possible to prevent all potential disaster scenarios, but you can take advantage of numerous redundant and recovery capabilities in the cloud to minimize any downtime caused by disk or hardware problems, power outages, or human error.


With cloud computing infrastructure, you gain the flexibility to choose the best technologies in software, platforms and computing resources.


Cloud computing infrastructure can grow or shrink easily and quickly. If your business experiences a burst of traffic you can quickly scale upwards to meet the demand. On premise you might be waiting for a shipment to arrive before you can scale!


Cloud providers can provide security through scale and standardization across their cloud computing infrastructure. 

Limitations of cloud infrastructure

Migrating to the cloud can have some limitations that are not always obvious. Let’s review some of the main disadvantages of using cloud computing infrastructure:

Limited control 

Utilizing a managed set of cloud computing services is ideal for scalability and cost savings, but that comes at a price in the form of control. Since cloud computing infrastructure is hosted by the cloud service provider, customers will have limited control over the administrative controls and no direct access to the backend infrastructure. Typically, this is not an issue for most companies, but something that should be considered.

Dependant on internet connectivity

Many of us take internet access for granted, but a network outage or slow internet speed will affect the quality of the services from the cloud provider. If you experience a loss of internet access you will be unable to take advantage of any of the hosted resources.  


Yes, this was listed as an advantage, but can also be a limitation if not properly configured.  When you deploy workloads to the cloud, you are relying on the cloud provider to provide overall security of the platform, but you are still responsible for securing your own applications, implementing best practices, and enforcing cloud governance policies.

Types of cloud computing services

According to Gartner’s latest report, public cloud end-user spending is expected to exceed 480 Billion in 2022. Let’s take a look at the three most common cloud computing layers offered on public cloud computing infrastructure:

worldwide public cloud services spending

Software as a service (SaaS)

SaaS is the most commonly used type of cloud service and is delivered to the end user as a complete application. The operating system and underlying cloud computing infrastructure used to host the software is hidden from the user. Familiar examples of a SaaS solution include Dropbox, Cisco WebEx, Salesforce, and Google G Suite. 

Platform as service (PaaS)

PaaS is a set of services that provide a framework for creating, running, and managing business applications. This service is popular with developers since it allows them to focus on writing code rather than managing the runtime environment, operating systems, and middleware. Microsoft Azure, OpenShift, and AWS Elastic Beanstalk are common examples.

Infrastructure as a service (IaaS) 

Iaas is an aggregate of cloud computing infrastructure in the form of virtualized computing resources. Rather than purchasing and maintaining hardware on-premises, companies can order virtual servers, databases, storage and networking services on-demand. Amazon Web Services (AWS), Google Compute Engine (GCE), and IBM Cloud are just a few examples. 

Public, private, hybrid and multi-cloud deployments

Cloud computing infrastructure components remain the same across the four common cloud deployment models. Each delivers cloud computing shared resources to the consumer through virtualization of the physical hardware, storage and networking. Let’s take a look at what makes them different. 

Public cloud

Public clouds are managed by a 3rd party and are available over the internet. Their resources are self-service and made available on-demand; this allows consumers to scale without the upfront cost of maintaining their own cloud computing infrastructure. Customers are accountable for securing the applications they deploy while the cloud provider is responsible for securing the cloud computing infrastructure. There are plenty of public cloud options, and most offer free cloud computing resources to get started. This is an ideal option for customers that are looking for a low-maintenance, low-cost, and a scalable solution.

When migrating to a public cloud deployment model you are not limited to a single cloud computing infrastructure provider. If a company wants to take advantage of industry leading cloud technologies, using multiple public cloud providers (multi-cloud) provides some flexibility in choice and could potentially reduce cost if applications are containerized and portable.  

Private cloud 

In a private cloud deployment, the cloud computing infrastructure can either be deployed on-premise or 3rd party hosted; dedicating access in either case to a single company. Companies have full control on where data is stored and how the networks will be routed. This is ideal for companies that want to maintain the security of an isolated environment and have more granular customization control while still providing the advantages of cloud computing.  This additional control has its disadvantages since it could be more expensive and may not be as scalable as public cloud.   

Hybrid cloud  

A hybrid cloud computing architecture utilizes cloud computing infrastructure from both a private and public cloud. The flexibility of this deployment model allows the company to decide which applications and services should be kept within their firewall for maximum security and workloads can take advantage of the cost savings of public cloud computing. Companies that are considering a full move to the public cloud will often implement a hybrid cloud solution initially.  

Wrapping up cloud infrastructure

To summarize, cloud computing infrastructure are the components necessary to implement cloud computing. Technology breakthroughs over the last 70 years have made it possible for on-demand resources to be shared over the internet. With a variety of deployment and cloud services to choose from, low upfront costs, and industry leading technologies available it is not a question of “if” a company will eventually adopt a cloud infrastructure but “when”.  

I hope this post was helpful in understanding a little more about cloud computing infrastructure and how it can be used in your journey to the cloud. Thank you for reading.

John Conrad, Solutions Architect, Capital One Software

John Conrad is a Solutions Architect at Capital One. He has worked with containers and Kubernetes since 2016 where he helped deliver a K8s workshop for 300 partners. Prior to joining Capital One, John was a WorldWide Technical Sales Leader for IBM Collaboration Solutions where he had the privilege to travel to 20+ countries. John has 25 years experience in software and technical sales and holds a Computer Science degree from the University of Kentucky. You can connect with him on LinkedIn (

Related Content