What’s Cloud-/Edge-Computing? | A Beginners’ Guide

14 min readMar 24, 2021

Already today, companies start managing their IT resources via the cloud. Smart apps are centralized. The edge will become a data filter.

Two years ago, the market research company IDC estimated the amount of data collected by mankind to date at 18 zettabytes (ZB). One zettabyte is equivalent to just over 1 trillion gigabytes. Today, in 2020, this number is likely to have more than doubled. Looking five more years into the future, experts forecast the expected amount of data to be as much as 175 zettabytes. If all this data were to be burned onto DVDs, the pile would circle the earth 222 times. But this is only the beginning. All new technologies, including in particular the Internet of Things, deliver a large amount of data every second. With digitization and more and more everyday applications, we are facing exponential data growth.

“I don’t need a hard disk in my computer if I can get to the server faster… carrying around these non-connected computers is byzantine by comparison.”
- Steve Jobs

But the data we collect must also be stored somewhere, almost more importantly managed. This consistently increases the demand for IT resources. Cloud and edge computing provide a possible answer for a sustainable future model.

Summary: The Cloud as an administrator of our data?

Nothing “new”. Already with the development of the ARPANET, the first primitive version of the internet in the early 70s, work is being done on models of the cloud.
Private users know cloud based applications mostly in the form of Gmail, Office 365 and Instagram.
In simple terms, cloud computing means that companies move their IT resources, such as databases, servers and storage, to the Internet (the cloud).
Well-known cloud computing providers are Amazon Web Services (AWS), Microsoft Azure and Google Cloud. But also Salesforce, Alibaba and other large companies have their own cloud services.
There are public, private and hybrid clouds as well as different types of services like Paas or SaaS.
Edge computing is an intermediate layer between the application and the cloud/data center, which only forwards part of the data. This is used especially for Internet of Things applications to create a new effectiveness.

How does cloud/edge computing work?

Cloud computing is the on-demand delivery of IT resources over the Internet. Storage, servers, databases and software are no longer purchased expensively by individual companies and managed independently. Instead, they are accessible via the Internet (the cloud) and rented via Amazon, Microsoft, Google & Co. Among other things, this enables real-time scaling with massive cost savings. While your own hardware may be left unused, cloud services are only rented for what is actually needed. If demand increases massively — for example, during Black Friday or Singles Day at online retailers — the system grows with it. If the systems are used less, less is rented.

According to a study by Data Group, 69 percent of all companies already use cloud computing in some form or another. However, only 20 percent of workflows take place in the cloud. Accordingly, the relocation of complete work processes to a virtual environment is not yet standard practice. In everyday life, the cloud meets private users mostly through the storage space in dropbox, OneDrive & Co.

Overview Top 3 Cloud Types

But not all clouds are the same. There are three different types, which are in use worldwide today. All have their advantages and disadvantages. That is why I would like to name them briefly.

Public Clouds: These are owned by external providers and are accessible online. The services are offered via the Internet and, unlike the private cloud, are not only accessible to individual companies. The service models are usually based on performance, but there are also subscription models. Examples of public clouds are Microsoft Azure, Amazon Web Services or the Google Cloud.
Private Clouds: Are in the hands of individual companies. Their own services and infrastructure are managed in a private network and are accessed either via a separate intranet or a virtual private network (VPN). This enables a high level of control and security.
Hybrid Clouds: Allow data and applications to move between private and public clouds. This becomes interesting when security factors are in the foreground. While non-critical data can be managed via a public cloud, security-critical information is managed in a private cloud. However, operating a hybrid cloud also requires the most effort of the three models.

What is the Cloud Computing Stack?

In addition to the cloud types, there are also a variety of service models, the Cloud Computing Stack, that companies can use in public clouds.

IaaS (Infrastructure-as-a-Service): is the simplest category. Only the IT infrastructure (storage, networks, etc.) is rented, just as most companies have in their data centers.
PaaS (Platform-as-a-Service): As its name suggests, Platform-as-a-Service models provide companies with an environment for developing and testing software. It thus builds on IaaS.
SaaS (Software-as-a-Service): SaaS applications have become very popular in recent years due to their easy access via the Internet. Microsoft Office 365 and Google’s G Suite are undoubtedly among the most popular. The service model first used by Salesforce back in 1999, software applications and infrastructure are hosted by the providers. These take over maintenance work and administration.

Edge computing in times of IoT

There are time differences when transferring data between its creation and management in the cloud. The so-called latency. This is a critical factor for IoT applications that rely on real-time analysis of thousands of safety-critical sensor data — such as autonomous vehicles or aircraft. Imagine a blind, self-propelled car.

This is where edge computing comes in. The edge, the so-called edge of the network, forms a data filtering layer between the sensors and the data centers. Its goal is to minimize latency, prevent network congestion and ensure functionality.

This is done by evaluating the data where it is collected. In this way, the Edge filters the security-relevant data that the system requires for its functionality and only forwards the collected data to the cloud that is relevant in the long term. More information about Edge is available in the next articles.

Cloud applications are on the rise.

The cloud enables companies to flexibly rent IT resources over the Internet. Many technology companies can hardly manage without today’s Software-as-a-Service products. Especially in the age of COVID-19, Office 365 or Zoom, to name just two, have become part of many people’s standard repertoire. Together with the processing of data at the edge of the network (called edge computing), this forms the basis for many future applications.

Above all, the Internet of Things and artificial intelligence require cloud and edge computing as a basis. All their applications depend on the analysis of large amounts of data, because smart cities and smart devices communicate a large amount of collected data every second. Via the clouds, companies can analyze and summarize this data and use it for further applications.

Cloud computing thus requires a good and stable Internet connection. However, this is usually still difficult, not only in rural areas. This is where edge computing comes into play. It serves as a filter between applications and the cloud. IoT devices generate a lot of so-called “disposable data” that is needed in real time for decision-making, but not afterwards. For example, an autonomous car collects up to 4 terabytes of data per day. An airplane collects 150 million data per flight. It is estimated that the Edge can manage 43 percent of this IoT data. This means that the cloud only receives the data it needs for further processing.

Advantages & disadvantages of the cloud

A stable Internet connection is probably the greatest weakness of cloud computing. But the security of the cloud should also be noted. With public clouds such as Amazon Web Services (AWS) or Microsoft Azure, this depends largely on the provider. However, it is questionable whether a medium-sized company can offer more security than these industry giants. Large companies have the option of a private or hybrid cloud to manage security-critical data even more effectively.

The advantages of the cloud are primarily the simple scaling. If IT resources are rented, they can be increased or decreased within a short period of time, even if required. This is a particular advantage for seasonal business, such as Christmas trading. At the same time, most companies are likely to record cost savings. The service costs at the cloud provider replace the investment costs for own data centers, hardware and software as well as daily operations. Unused resources are a thing of the past. As a result, companies can focus on their business again and not take over any surplus administrative tasks. By connecting to a cloud, companies can also more easily access additional modules. Thus, the large clouds all offer solutions for AI and machine learning or even intelligent analyses.

Cloud-based applications

Many of our current cloud-based applications are Software-as-a-Service (SaaS) models. They are hosted in clouds and accessible via browsers and smartphones. According to Statista, the SaaS market has grown from nearly $5.56 billion to over $157 billion since 2008. What a development!

“Imagine if we can get information from sensors, compute it in the cloud, and find and plug water leakages — leakage wastes 20 percent water in the world.”
Rajeev Suri, CEO of Nokia

Internet companies

Trello, Atlassian and other well-known Internet companies have shifted a large part of their IT structure to the cloud. No wonder. Many of their processes and their customer solutions are SaaS models and therefore accessible to their customers via the browser. Salesforce, one of the founders of SaaS, and other companies offer their e-commerce solutions with ever new applications via the cloud. Their success seems to prove the model right.

In the future, many of these Internet companies are likely to be added to the Internet of Things.

Media

Netflix, Amazon Prime Video and other streaming service providers rely on cloud-based IT solutions. By 2017, more than 125 million hours of video content were streamed daily on Netflix alone. This number is likely to be significantly higher today. The IT systems not only have to be correspondingly stable, but also scale with the hourly changing numbers of streamers.

Furthermore, Netflix is a strong supplier of its own content. The American company has collected 430 nominations and 72 awards (29 of which for House of Cards) for its own productions in recent years. The content produced depends to a large extent on user behavior. In order to find the most popular genres, actors and directors, Netflix relies on extreme data analysis, which is also likely to require a large amount of IT resources. Incidentally, 80 percent of the titles viewed were suggested to customers by Netflix’s own $1 billion algorithm, whose calculations are also performed in the cloud.

Retail

Many customer relationship management (CRM) applications are only available to companies in the cloud as software-as-a-service models. But retail companies are also increasingly turning to cloud-based microservices for their own architecture. Their e-commerce platforms run on compute engine instances. Complex models with high computing power, e.g. for personalized product recommendations, are thus much easier to scale and adapt to business conditions than was previously the case. The CPU-intensive processes can be executed in the cloud, just like the increasingly important data analysis. This makes the Omni-Channel experience available not only to customers but also to employees, who can more easily access a wide range of applications.

Healthcare

If you google “Healthcare Cloud”, you quickly come across special cloud applications from Google, Microsoft, Salesforce, etc. The providers are positioning themselves because the healthcare system of the future is no longer decentralized. Data will be stored and managed in the cloud, and doctors, patients and insurance companies will be able to access it. Seamless data management and exchange between the various parties enables powerful joint analyses. In combination with new applications such as Internet of Things devices (e.g., sensors that monitor our vital signs 24/7) or artificial intelligence that detects diseases in time through the values provided by the sensors, this will disrupt the industry. But cloud computing will also enable telemedical applications.

AWS, Netflix & Co: providers and users of cloud computing

Amazon, Microsoft and Google are driving the cloud. Cloud users range from Audi to Netflix to Siemens.

If the majority of companies were already aware that the future is digital, the COVID-19 pandemic may have convinced the last doubters. With the increased need for remote work, interest in converting business processes to the cloud is also growing. Not only the fact that employees can store their documents in cloud storage, but also the scalability of IT resources based on the unpredictability of today’s world is likely to be a major reason for the cloud.

Once companies have decided to move their IT resources to the Internet, they are already faced with the next question. Which provider is the right one? More and more technology leaders are trying to strategically position themselves in the market of the future. They usually differ in the number and quality of their own services for artificial intelligence, Internet of Things, edge computing and big data analysis. With their easy integration into the company clouds, expert tools are available to medium-sized companies at the push of a button. Years of research & development are no longer necessary.

Top 3 Cloud Provider

The cloud market is expected to be worth around USD 626 billion by 2023. According to the Magic Quadrant Report 2019 from the market research company Gartner, this is largely distributed among the top 3 providers Amazon, Microsoft and Google.

In contrast, China’s e-commerce giants Alibaba and Tencent, with their clouds, are still considered niche players. They are strongly represented in Asia, but are still largely unused in Europe and America, partly due to geopolitical pressure. Furthermore, according to Gartner, non-Asian customers do not have access to the full performance or functional depth of the clouds as is the case with the competition. Oracle and IBM are also considered specialized.

Amazon Web Services (AWS)

https://aws.amazon.com/de/solutions/case-studies/

Amazon has taken on a central role in our society in recent years. However, the world’s largest online retailer has also become the absolute leader in the cloud market. A full 58 percent of its operating profit comes from this market, which has thus become the company’s growth driver even before the retail trade. An important area for future positioning is artificial intelligence and IoT applications.

Microsoft Azure

https://azure.microsoft.com/de-de/case-studies/

Amazon’s largest cloud competitor is Microsoft. The Windows provider scores with its easy integration with services of its operating system. Its strength lies in the establishment of Azure in its platform concept, and for medium-sized companies, this is precisely one of the main reasons for their decision. Apart from this, they also focus on artificial intelligence, Internet of Things and other services to further simplify the adaptation of future technologies for their customers.

Google Cloud

https://cloud.google.com/customers#/

The Google Cloud platform is seen as a solid number 3. Although Google has been able to gain some larger customers and score points, especially through Google Meet as part of the G-Suite, the cloud is not yet so widespread and still lacks some features compared to its major competitors. However, in terms of pricing, Google has an advantage.

Audi, Netflix, Ubisoft & Siemens: Users of the Cloud

If you look at the case studies of the top 3 cloud providers, you will find a lot of well-known names. In the following, I have picked out four exciting examples and examined them a little more closely.

Audi

The Bavarian car manufacturer uses cloud computing for simulations in autonomous driving. Up to five cameras are already installed in Audi’s. In combination with radar, ultrasonic sensors and navigation systems, up to four gigabytes of data will be generated in the future. Mind you, per second. Most of this data is needed for real-time processing and is managed in edge computing.

If you want to simulate an autonomous car drive, as in Audi’s case, you need all these data. The simulations are essential for the development of new software. Audi relies on a cloud platform for the use of a virtual environment and the application of raw sensor data in different situations. This allows them to scale with IT performance. According to Microsoft, a cluster with approximately 12 petabytes of storage capacity and 60 computing machines was used in parallel.

This allows Audi’s vehicles to undergo all kinds of tests with the latest software before it is rolled out to customers worldwide.

Netflix

Streaming service provider Netflix has turned most people’s television viewing habits upside down. With 167 million subscribers in more than 190 countries, the company is facing major technical challenges. As early as 2017, more than 125 million hours were streamed daily.

The on-demand availability of video content allows customers to choose when they want to consume something. For Netflix, however, this makes it more difficult to predict access and IT resources required. Other influences, such as the 17 million new registrations during the first months of the COVID-19 pandemic, make realistic planning difficult beyond that.

To respond to the new challenges, the IT structure of Netflix 2016 was completely migrated to the cloud. Since then, they have been one of the largest AWS customers. It is estimated that Netflix cloud computing costs are $9.6 million per month.

Ubisoft

Assassins Creed, Watch Dogs and Far Cry are just a few of the games that gamers around the world are likely to know. The French game developer is one of the most famous faces of the scene and tens of thousands of players use the online multiplayer modes.

Especially during traffic peaks on holidays or weekends, the company depends on fast and easy scalable multiplayer gaming. In the case of Watch Dogs 2, up to 120 relay servers were used and 70 terabytes of data were managed in 20 days.

In addition, the Ubisoft cloud enables faster delivery of updates.

Siemens

The German technology company Siemens uses artificial intelligence for cyber security in its Cyber Defense Center (CDC). The algorithms, which are controlled via Amazon SageMaker, evaluate 60,000 threats per second and require only 12 employees in administration. At the same time, historical data can be evaluated.

“Our goal was to use cloud-based artificial intelligence to process these huge amounts of data and make immediate decisions about how best to counter any detected threats”
Jan Pospisil, Senior Data Scientist am CDC

This enables Siemens to react quickly to new threats and learn from old ones. In times when many things change quickly, this seems to be a good strategy.

Cool Stuff

Ever heard of IBM’s cloud-based chemistry lab called RoboRXN? If not, then it’s about time! It’s the perfect example of how exponential technologies will work together in the future. In the case of RoboRXN, the cloud is linked to artificial intelligence and automation (robotics). Scientists can access an application online where they can create their chemical formulas with the help of AI. The manufacturing process is then automated. The future is within reach: https://rxn.res.ibm.com/rxn/robo-rxn/welcome

During my research on the cloud, I repeatedly stumbled across CloudComputing Insider magazine, which helped me a lot with its definitions of the individual terms. Absolute reading recommendation: https://www.cloudcomputing-insider.de/

Sources