Edge Computing

Edge Computing: On the edge and yet right in the middle

"Edge" is one of THE buzzwords in the IT world. In the age of the "Internet of Things" or "autonomous driving" as well as other modern applications, edge computing has mutated into the superstar of digitalization. Almost everything must be "edge-capable" to meet the demands of the latest technological achievements.

There are different interpretations for the idiom "edge". Those who stand on the edge are still part of the whole - just not in the middle. This has advantages and disadvantages: Those who are on the edge can sometimes be cut off from important information or receive it late. However, those who are not in the middle of things sometimes have a clearer view of the big picture and can provide important impulses for improvement. On the periphery, IT processes can also be handled without the bulk of the remaining processes. In some areas this changes entire process chains.

Edge Data Center – old ideas in new packages?

How does the term "edge" relate to the data center? First of all, it should be made clear that an edge data center is not fundamentally different from a classic data center in its structure. The only difference is that it is outsourced, installed at the edge and not in the middle of the network.

Edge data centers are predestined for local data processing, for time-critical or bandwidth-intensive applications. Market researchers unanimously forecast high growth rates for edge computing in the coming years. The driving force behind this is the dynamically increasing number of IoT applications. In addition, there are more and more on-demand services as well as autonomous traffic and transportation systems.

As many large cities around the world are on the verge of traffic collapse, new mobility concepts are in demand in urban areas. The key to the successful realization of such applications is 5G mobile communications. The high bids for the corresponding licenses are an indication of the growth that can be expected in these segments.

The high license costs for 5G are one aspect of this development. In addition, however, there is also the need to invest massively in new network infrastructures. Especially companies with grown infrastructure are still quite hesitant in this respect. They are afraid of high investments, which would require a complete rebuilding of their network infrastructure. But those who do not support and implement changes may lose touch with prospering markets. A dilemma.

Where is edge computing used?

So there is no doubt that new infrastructures will be necessary. Not least the multitude of new applications, which are rapidly establishing themselves in individual industries, require quick action here. Logistics and industrial production, for example, are confronted with a massive wave of automation. More and more sensors and an increasing number of new components and participants are coming into play. They generate and process data and thus provide important information for numerous processes.

Other factors such as the safety of production processes and machines should also be taken into account. Even in the event of power failures or other malfunctions in the data center, uninterrupted operation of all components must be ensured in order to guarantee smooth processes. Some of these tasks can be solved by using artificial intelligence (AI). This applies to robot control as well as to quality control or component identification, to name just a few fields of application.

Most of these functions require local computing power. However, artificial intelligence and "machine learning" also require access to learned patterns of identical or at least similar machines. This simplifies problem recognition and enables proactive action. The combination of edge computing and cloud computing is perfectly suited for such applications.

Combination of Edge and Cloud supports SMEs

The combination of edge computing and cloud computing is also an attractive option for small and medium-sized enterprises (SMEs). More and more applications such as classic office packages, collaboration solutions as well as ERP and CRM systems are now available in the cloud. Investments in complex IT infrastructures are therefore no longer worthwhile for many SMEs. The cloud is economically more attractive. Nevertheless, there are still a number of applications whose data must be processed locally for security reasons. With edge computing, the necessary local infrastructures can be easily set up. Communication via cloud enables seamless processing of all relevant information in any business application. The combination of Edge and Cloud offers SMEs completely new methods for opening up new markets and strengthening competitive positions.

Mobile applications drive Edge growth

However, the most rapid increase in edge applications is likely to be in the area of mobile applications. A crucial point here is the growing number of IoT devices. Services such as SigFox, NarrowBand IoT, WiFi6 and 5G will significantly increase the data volume. The real-time requirements for data processing will also increase.

Here, a division of services, the so-called slicing, can bring relief. This procedure allows different treatment of applications, for example locally for real-time processing, and it allows traffic consolidation. Less critical data then migrates to a centrally stored cloud application for processing.

The cost issue is also a key factor in this environment. The construction and expansion of networks requires considerable investment. This can easily prove to be an obstacle to further developments and the nationwide distribution of modern services. For example, 5G expansion requires significantly more sites, as much smaller radio cells are used. This in turn makes it necessary to provide suitable network connections for the base stations. The network topology will thus be broader overall and additionally meshed in width. Edge systems can be established at the nodes of this wide meshed network to ensure the seamless operation of applications.

Trend researchers see Edge as gaining momentum

Why is Edge an issue right now? This is probably primarily due to the change that is taking place due to completely new applications and the urge for mobility.

In the past, after a consolidation phase, the necessary compute power moved closer and closer to the clients. Today the number of local workstations is decreasing. Mobile devices are increasingly taking their place for numerous applications.

Current studies show a clear trend towards the use of cloud applications in companies. The hype is over, cloud computing is standard almost everywhere. Especially in the public cloud sector, there are substantial leaps in sales. A recent study by ActualTechMedia shows that public cloud customers see considerable benefits in almost all important areas. These primarily include adaptability, scalability, availability, security and more attractive operating costs. According to this, edge computing is an optimal complement to the public cloud and scores with these points:

  • Flexibility
  • Scalability
  • Availability
  • Efficiency and ease of handling
  • Security

According to network specialists, the development of edge computing is still in its infancy. The potential and the number of possible applications are high - on the edge and yet right in the middle.

Do you have any experience with edge computing so far? What opportunities do you see for edge computing in the future?