Read on:
Virtualization Trends Series: A Brief History of Virtualization: Part 1
Virtualization Trends Series: The Evolution and Future of Hypervisors: Part 2
Virtualization Trends Series: Do you Actually Need a Multi-Cloud Strategy: Part 3
Virtualization Trends Series: ChatGPT and PowerShell: Part 4
Growing need for Edge Computing
In this part 5 of our Virtualization Trends series, we will discuss Edge computing. In case you are not too familiar with it, Edge computing is often used in Cloud Computing scenarios and refers to moving data processing to the devices producing it as opposed to shifting it over the wire to the cloud instances. This offers the benefits of saving on data transfer costs and time as well as catering for those cases where the edge device may not always have internet access available.
The IDC (International Data Corporation) describes edge computing as a meshed network of micro-datacenters, capable of storing and computing data locally. Edge computing mostly applies to the Internet-of-Things (IOT), but it also addresses several use cases that will only keep growing in number. Back in 2018 already, Gartner estimated that 75% of data processing will be done outside of the traditional datacenter or cloud by 2025.
5G as a driver for Edge Computing
Industry experts consider the deployment of 5G antennas to drastically improve coverage, as a significant driver for organizations willing to leverage Edge computing. 5G is extremely fast with low latency in order to cater for the ever-increasing amount of data being produced by user devices. While this is great, this only applies to the connection of the device to the antenna. This network infrastructure in the backend that needs to convey the data to its end location (cloud services, facebook, you name it) isn’t evolving as fast as the deployment of 5G towers.
This causes a bottleneck that could be mitigated by moving some of the processing off the centralized cloud location and distributing it closer to the user, thereby avoiding the need to send the data down the pipe to get a response faster and fully leverage the capabilities of 5G networks.
Another use case for this is that antennas would be prime locations to host micro-datacenters which could be rented by nearby professionals who cannot invest in their own datacenters.
IDC forecast increased spending in Edge Computing
Back in early 2022, IDC released their predictions as part of their Worldwide Edge Spending Guide in which they forecast a sustained growth throughout 2025 with spendings that could reach up-to $274 billion. It is said that most of the spending will be on hardware making up 85%.
Enterprises are expected to heavily invest in 6 key areas:
- Manufacturing operations
- Production asset management
- Smart Grids
- Omni-channel operations
- Public safety & emergency response
- Freight monitoring and transportation systems
Still according to the IDC, the main use cases for Edge Computing will be Business intelligence/data analytics, Content delivery, Text and image analytics as well as Networking and Security.
Leading Edge computing companies
Hundreds of companies are heavily involved in the development of Edge Computing. Obviously, it is quite tricky to make a list of top companies as they don’t all operate in the same areas such as platform, system, hardware, datacenters, processors, cloud services extensions and so on. As a result, we will mention only a few that we should keep an eye on.
You won’t be surprised to find the usual big players among the industry leaders but there are other less known company names that come up in this space too.
- AWS: AWS edge solutions extend their cloud expertise to hybrid cloud scenarios with AWS Outposts, AWS Wavelength and AWS Local Zones. You might also be surprised to learn that they offer edge devices (hardware) with AWS Snow
- Microsoft Azure: A set of services geared towards edge computing is proposed as Azure private multi-access edge compute or Azure MEC. It is made of partnerships and Azure services like Azure Stack Edge or Azure IoT
- Dell: The company offer a wide range of edge devices aimed at tackling enterprise edge computing needs with servers, storage or networking equipment
- EdgeConneX: This company is a datacenter provider that builds and operates over 40 datacentres around the world. They also create purpose-built datacenters with their Edge Data Centers and Far Edge Data Centers for those customers that require specific or ultra-low latency capabilities
- And more…
Containers and Serverless
The two main approaches to Edge Computing are to leverage containers or serverless when it comes to running workloads. We aren’t discussing virtual machines (VMs) here as, while some companies like VMware offer Edge computing solutions based on them, the overhead involved with virtualization makes it a bit difficult to consider it as a solid contender.
Containers, usually built around Docker or Kubernetes offer great portability as you can deploy the same code to different locations. It is vendor agnostic, offers great version control and solid isolation. However, the complexity associated with running containers, especially with Kubernetes, can sometimes put a downer on the prospect.
On the other hand, serverless offers a different approach in order to make developer experience simpler by reducing the overhead of running and managing code. They work similarly to Functions as a Service (FaaS) like AWS Lambda, Azure Functions and Google Cloud Functions. At the Edge, you can find, among others, Akamai EdgeWorkers which lets you deploy JavaScript functions at the edge. Serverless will be cheaper, potentially more flexible and less complicated. However, it does mean accepting vendor lock-in and might complicate building complex apps.
Conclusion
By placing data processing closer to the devices producing and consuming it, companies delivering Edge Computing capabilities offer game changing benefits like faster insights, lower latencies and increased bandwidth performances. As a result, many companies are moving processing from their datacenters or cloud services to the edge.
This growing adoption will also hopefully relieve WAN networks that have been sustaining a constant and incredible growth in data being transferred over the last decade (data decade).
Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.
