3 Best Cloud Computing Services (AWS/GCP/Azure) In 2022 [How to Configure + Real-World Examples]

What is Cloud Computing?

The NIST (National Institute of Standard and Technology) defines cloud computing as, “a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.”

This definition is summarised as “an on demand model that gives room for sharing a pool of resources which are: Servers, Storage, Applications, Services, etc”.

According to Wikipedia, “Cloud computing is the on-demand availability of computer system resources, especially data storage (cloud storage) and computing power, without direct active management by the user.”

NIST cloud computing model covers both the deployment model and service model, in this case, NIST categorised deployment model to contain four basic types of model which are public, private, community, and hybrid deployment models. While the service or delivery model is categorised into three fundamental types of model which are; 

• Infrastructure as service (IaaS)
• Software as a Service (SaaS)
• Platform as a Service (PaaS)

Benefits of cloud computing services

• Cloud computing has made business very vast, by reducing stress on clients through proper signal coverage; every client or user can access the cloud hosted application irrespective of location as long as they have access to the internet.
• Cloud computing provides flexibility in its service, this means flexible infrastructure can be turned down, up, or off based on the users’ condition or state.
• Affordability of Cloud computing has proven to be very cost effective as it takes away the excessive cost of running a desktop application, power, and other maintenance. Storing and processing your data on cloud systems will only cost you a one-time payment or pay as you go.
• Cloud computing operates 24/7 and this makes the uptime very high. According to idexel.com, the uptime of almost all cloud vendors is rated to be 99.9%.
• Automatic software update in cloud computing, all the stress of maintenance of both hardware and software is taken care of by cloud service providers, and no extra charge is asked of the clients to pay for maintenance.

What are the main types of cloud computing services?

The main cloud computing services are: 

• Infrastructure as a Service (IaaS)
• Software as a Service (SaaS)
• Platform as a Service (PaaS)
• Function as a Service (FaaS)

      …

1. In Infrastructure as a Service (IaaS) model, services such as storage space, database management, and servers are offered based on pay per use to consumers. Examples of IaaS are; Amazon Web Services, GoGrid, etc. In this model, the admin has total control of the hardware while consumers have total control of the applications, middleware, and guest operating system.

2. In Platform as a Service (PaaS), consumers have no control over operating system and hardware but the service provider has total control over both operating system and hardware. Though the consumer has admin control over the application and also has access to interface his Program. Some benefits of PaaS includes; saving cost upfront, use of software licence efficiently, central management of data, etc.

3. In Software as a Service (SaaS), scalable applications that are internet based are deployed to the cloud to offer services to internet clients. Examples are; Google Docs, saleforce.com, etc. Consumers have no control over the hardware, operating system, and middleware, but the cloud service provider has complete control. Consumers have only minimal admin control over applications and user level management. Some of the benefits of SaaS are the same, such as upfront cost savings, effective use of licences, central data management, and so on.

4. Function as a Service (FaaS) is a type of cloud computing service  built around serverless computing. The aim is to isolate the hassle of deploying and maintaining cloud infrastructures before running the written program; instead, the program is uploaded to the cloud to run based on some event that triggers the program.

Figure 1 – Overview of Cloud Computing Services

3 Best Cloud Computing Service Providers for 2022

According to the survey result from StackOverflow, the top 3 most popular cloud computing companies in 2021 were AWS, Microsoft Azure, and Google Cloud taking the position of 1, 2, and 3 respectively. Therefore, we will describe them one after the other. 

1.  Amazon Web Services (AWS)

AWS (Amazon  Web Services) is an online platform that provides scalable and cost-effective cloud computing solutions. As the world’s leading cloud service provider, AWS is a comprehensive and broadly adopted cloud platform. AWS offers services like:

• Infrastructure as a service (IaaS)
• Platform as a service (PaaS)
• Software as a service (SaaS)

AWS provides over 200 fully featured services from its data centers worldwide (growing exponentially). Provides on-demand service delivery with pay-as-you-go pricing, which is one of the key reasons why it is appealing to millions of clients, including numerous startups, big corporations, nonprofit organisations, and government agencies. AWS is also cost-effective, scalable, and secure.

What is AWS used for?

AWS gives more services and features to its customers than any other cloud provider, this makes it faster and easier to build applications on the cloud. There are various uses of AWS like:

• Data Lakes
• Analytics
• Machine Learning
• Internet of Things
• Containers
• Serverless Computing
• Storage, and a lot more.

Figure 2 – Key AWS Features

AWS has various features which make it reliable among different firms. Some of the key features are:

Serverless Computing

AWS Serverless Cloud Service assists its clients in such a way that they may concentrate solely on developing applications. AWS manages the servers in addition to scaling, patching, and administration of the infrastructure.

AWS Lambda 

A serverless, event-driven compute service that lets its customers run code for virtually any type of application or backend service without provisioning or managing servers. Lambda can be triggered from over 200 AWS services and software as a service (SaaS) applications, and only pay for what the client uses.

AWS Cloud Databases

Amazon databases are highly performant, completely managed, and cost-effective databases for customers. AWS provides the broadest selection of commercial-grade purpose-built databases, allowing the development team to build and innovate faster on any application use case.

AWS offers variety to database engines, which includes:

• Relational Database – Transactional Purposes.
• Non-Relational Database – Internet-Scale Applications.
• Data Warehouse – Analytics.
• In-memory data Store – Caching & Real-time workloads.
• Graph Database – Application with highly connected data.

Cloud Storage on AWS

Cloud Storage on AWS provides speed, avoids complex capacity planning, and reduces over-provisioning with ease. The AWS storage can be used independently as well as in combination to meet customers’ requirements. Types of storage provided by AWS are:

• Amazon Glacier – Used for long-term storage.
• Amazon Simple Storage Service – Provides scalable object storage for archival, analytics, and data backup.
• The Amazon EBS – Provides block-level storage volumes for persistent data storage for use with EC2 instances

How does AWS work?

AWS takes care of the heavy lifting for its customers, so they can build and run virtually any type of application, including social networks, climate change research, gaming, video streaming, online banking, and many others. These services are used without upfront costs or long-term commitments.

Thanks to the pay-as-you-go model, customers are available to experiment with different technologies until they find ones that fit their needs. That way, startups finish building their product and features while keeping their cost minimal.

Some AWS services give full control to the customers while AWS manages the infrastructure and underlying resources. For other services, AWS is responsible for the operation and security of the entire stack so all that is required for customers is to use it. This is called the shared responsibility model.

When using AWS services, customers get to decide how much of the heavy lifting (if any) they would like to take on themselves.

Benefits of AWS

Figure 3 – Benefits of using AWS

Companies and individuals prefer AWS as their cloud provider because of the numerous AWS benefits it provides. Here are some of the reasons why:

User-friendly

AWS is easy to use as the platform is specially designed for quick and secure access. Users can modify their data whenever and wherever they want. AWS provides all the information, documentation, and video instructions to help users learn how to use all of its services.

Secure

AWS is architected to be the most secure and flexible environment available today as many companies rely on more secure infrastructures. AWS allows automating manual security tasks so the customers can shift their focus to scaling and innovating business. AWS follow different layers of data surveillance such as:

• Data protection
• Identity and access management
• Infrastructure protection
• Threat detection and continuous monitoring
• Compliance and data privacy

Flexible

Flexibility is also the reason why many companies prefer AWS. With a service like AWS EC2, customers can build their virtual computing environment by setting up their preferable operating systems and applications. AWS provides all the services that any application requires to function seamlessly. It also eases the migration process, and customers can work on new solutions simultaneously.

AWS Cost

When it comes to the AWS cloud, being knowledgeable about the pricing models and how best to optimize expense optimization is essential for avoiding surprises at the end of the month. With AWS customers only pay for the individual services they need, for as long as they use them. These prices can be calculated by different methods, but the main one is for the time of usage. Hours are the main factor in this method, followed by the minutes and seconds of use of certain services. The prices vary depending on the services and on the location of the resources being used by the customer.

The services we are always going to see on our AWS bill will be basic ones:

Storage

• EBS – $0.0010 GB per hour
• S3 – $0.023 GB per month

Computing power

• EC2 (Price depends on the server type, starting from $5 USD to over $150 USD x month)
• Load Balancer ($15 USD x 750 hrs)
• Elastic IP

Networking

• Inbound traffic – It’s free
• Outbound traffic – Starting with $0.09 GB < 10TB

AWS applied in the real world: Predicting Malaria Breeding Sites

One interesting case study based on AWS is Prediction of malaria mosquitoes breeding sites through satellite imagery and topography data can be found below.

Using the benefits of AWS s3 bucket and applying Deep Learning methods for storing, preprocessing, and classifiers training on satellite images, topography, and population density data of Malaria breeding sites in order to prevent the spread of malaria.

AI for Malaria Prevention: Identifying Water Bodies Through Satellite Imagery

2. Google Cloud Platform (GCP)

What is Google Cloud Platform used for?

Google Cloud Platform (further mentioned as GCP) is a suite of services offered by Google to companies and individuals. GCP was created in 2008 with the launch of ​​App Engine.  GCP services include everything from GCP cost management, to data management, to delivering web and video over the web, to AI and machine learning tools.

The services provided rely on: 

• Infrastructure as a Service (IaaS)
• Platform as a service (PaaS)
• Software as a service (SaaS)

What required companies to buy software and hardware, is now sold as a service.

Google services (GCP and other services) are served via 23 data centers worldwide based in North America, Latin America, Europe, and Asia.

Some of the top GCP customers are Paypal, Colgate-Palmolive,  eBay, Target, HSBC, Thales. In 2021, Google Cloud Computing captured 23% of cloud computing market shares.

How does GCP work?

There are 3 different mediums to manage projects and resources hosted by GCP:

• Google Cloud Console, a web-based, graphical user interface 
• Command Line Interface from your computer terminal using the gcloud command-line tool
• Client libraries based on APIs to access Google products (App APIs and Admin APIs)

Figure 4 – Google Cloud console

Key features of GCP

• Computing and hosting 
• Storage and database 
• Networking 
• Big Data 
• AI and ML

The photo below summarises key GCP services and features:

Figure 5 – GCP services and features

Benefits of Google Cloud Platform

Cloud Computing removes the financial burden of buying hardware and software necessary for certain activities by giving access to remote data centers owned by a third party. Cloud computing allows democratisation of digitalization, thanks to lower initial costs.

The key benefits of GCP:

• High Productivity with Innovation  
• Easy to Adopt Latest Functionalities
• Remote Access
• Smooth Collaboration
• Unmatched Security
• Reduced Breaches
• High Reliability
• Offers Flexibility and Control

GCP Cost

Cost is on a pay-as-you-use scheme. A detailed price list and price calculator are available on Google cloud website. Depending on services used, prices can be based on:

• Price per hour
• Price per month and per GB
• Price per node hour. 

Therefore, it might be difficult to know beforehand the costs you will incur. Luckily, as a first-time user, you can benefit from a 300 $ or 300 € free vouchers to use GCP. This voucher is valid for 3 months. 

A real-world example of the project realized with GCP

This case study shows how GEE was used in a real-world example in finding the best areas for solar installations in Africa with the use of nighttime satellite imagery.

Using Google Earth Engine for house roof detection through GEE available datasets and satellite imagery, in order to create a heatmap for solar panels locations in electricity demanding locations in Nigeria. GEE made it possible to investigate an effective image segmentation through connected components algorithm, work with GeoPandas and link the results with Google Maps for further exploration.

Satellite Imagery and Heatmaps to Identify the Best Areas for Solar Energy

3. Microsoft Azure

What is Microsoft Azure platform used for

Microsoft Azure is a cloud computing platform and online portal that allows customers (private clients and enterprises) to access and manage resources and services provided by Microsoft. It was launched in February, 2010 with the name of Windows Azure, being renamed Microsoft Azure in 2014.

How Microsoft Azure works

Microsoft Azure provides a range of cloud services, like compute, storage, networking, and analytics that allow users to develop, test, deploy, monitor, scale, and maintain applications by reducing (or eliminating) the hassle to plan, configure and maintain on-premise servers. 

Figure 6: Example of Microsoft Azure dashboard

Microsoft Azure holds 54 data centres spread in 140 countries around the world.  80% of the 500 worldwide most fortunate companies use MS Azure, with a market share of 30% amongst clouds. Some of the biggest customers include eBay, Boeing, Samsung, GE Healthcare and BMW.

Key features of Microsoft Azure

Microsoft Azure offers more than 120 services of the type:

• Infrastructure as a Service (IaaS)
• Platform as a Service (PaaS) and 
• Software as a Service (SaaS)

These services are divided into 18 categories. The most popular and used category services being: 

• Compute (Window and Linux VMs)
• Storage (like Disk and Blob storage)
• Containers (Azure Kubernetes Service), databases (SQL DB and CosmosDB), networking, analytics (Azure Data Factory)
• Mobile and web  
• AI + Machine Learning.

Figure 7: Most popular services offered by Microsoft Azure

Some benefits of Microsoft Azure include:

• On-demand scalability – which allows users to scale up their storage space and computing power on demand,
• No on-site hardware – which eliminates the need to configure and maintain hardware resources on-premise,
• Cost effective subscription models
• High availability – which increases the data redundancy across all data centres,
• Enterprise-level development tools – that allows users to develop their apps using a variety of popular tools, like the Azure’s support of multiple programming languages (C#, Node.js, Java),
• World-class cybersecurity – which guarantees that all data on Azure are protected by an advanced encryption process,
• Multi-cloud and hybrid cloud service.

Microsoft Azure Cost

Microsoft Azure offers different pricing models, convenient for any need. 

• Azure Free Tier: is the free contract that permits the free use of a variety of services for 12 months, with a limited allowance. In the free tier is included a credit of $200 to deduct from the first bill within the first 30 days.
• Spot Pricing: lets the user buy spare computing power at a discount of up to 90%. This service however could be interrupted by Azure on short notice, making it suitable only for workloads that can tolerate disruptions. 
• Azure Hybrid Benefit: this is for enterprises that already have a Microsoft server on-premise and want to build a hybrid solution.

Microsoft Azure as a base for projects

A case study of using computer vision for identifying malnutrition of children was based mainly on Microsoft Azure and here you can read how!

Increasing the accuracy of the neural networks’ algorithm identifying malnutrition of children by predicting height and weight of children through Computer Vision. The work was based on children scans (JPG, PCD, video, and depth maps), provided by Child Growth Monitor and conducted mainly with the use of Microsoft Azure.

Conclusion

AWS still dominates the cloud computing world and possesses cheap and reliable services compared to the other public vendors. However, Microsoft Azure and Google Cloud are catching up but not fast enough with AWS.

Furthermore, Omdena makes use of the three services based on stakeholders preference.

Author: Joseph Itopa Abubakar

Co-Authors: Alessandro Cubeddu, Muhammad Awon, Sandra Boniface, Patience Chizoba.

References

• https://aws.amazon.com/what-is-aws/
• https://aws.amazon.com/application-hosting/benefits/
• https://aws.amazon.com/pricing/
• https://www.simplilearn.com/tutorials/aws-tutorial/what-is-aws
• https://searchaws.techtarget.com/definition/Amazon-Web-Service
• https://k21academy.com/microsoft-azure/az-400/az-400-microsoft-azure-dashboard/
• https://www.saviantconsulting.com/blog/10-reasons-why-choose-microsoft-azure.aspx.
• https://marketplace.visualstudio.com/items?itemName=keesschollaart.AzureCostInsights.
• https://www.statista.com/topics/8048/google-cloud/#dossierKeyfigures
• https://www.bmc.com/blogs/aws-vs-azure-vs-google-cloud-platforms/
• https://www.acronis.com/en-us/articles/google-cloud-platform/
• https://searchcloudcomputing.techtarget.com/definition/Google-Cloud-Platform
• https://cloud.google.com/
• https://acloudguru.com/blog/engineering/what-is-google-cloud-platform-gcp
• https://www.c-sharpcorner.com/article/google-cloud-top-advantages-and-why-you-should-use-it-in-2020/