Acording to Forrester’s new report, if you’re confused about what “edge computing” really is, it could be due to marketing and hype.
According to ForresterNow’s report, The Four Edges of Edge Computing, there are four types of edge computing, but there is no hard, fast split between them. Also, the way vendors use this term makes it difficult to identify subtle differences.
“We like to draw clear boundaries as analysts, but in reality, real life is much more difficult,” said Brian Hopkins, vice president and report author of Forrester’s Emerging Tech Portfolio. I am. “Edges are best thought of as a broad category of related solutions that have something in common and attract different vendors, rather than strictly physically dividing the network. The four edges always describe the edge. It’s a client response that you have to do because different vendors use the term in different ways. ”
For example, collocation vendor Equinix uses the term edge to describe distributed enterprise workloads running in different colocation data centers, Hopkins said. As part of that, the Internet of Things (IoT) vendor Siemens uses the term edge to mean running software and analysis on servers in industrial manufacturing spaces.
“These two scenarios are very different and both use the term’edge’, but they are serviced by solutions from different vendors,” Hopkins said.
The report defines Edge as “infrastructure and software that is physically separated from, but connected to, enterprise core IT assets by shared fiber or wireless networks.” There are four types of edge computing outlined in the report. Each serves a specific niche: enterprise, operations, engagement, provider.
The Enterprise Edge consists of tier 2 and tier data centers, cloud gateways, colocation facilities, and office space connected by a wide area network (WAN). The goal of the enterprise edge is to enable customers to scale their core computing networks through enterprise infrastructure and WAN and public cloud vendors. Typical use cases include office automation, smart building, and e-commerce.
The operational edge consists of IoT networks and devices, local computing assets and gateways. These edge networks help automate on-site processes in extended locations such as retail stores, branches, manufacturing facilities, warehouses, storage facilities, and processing plants. Typical use cases include industrial automation, smart city features and services, logistics, and the realization of experience spaces.
Engagement edges are defined by globally distributed, high-performance computing clusters designed to improve customer engagement by reducing response times. These Points-of-Presence (POP) networks are physically close to the points of consumption to quickly connect customers to data and information. Typical use cases include streaming, gaming, zero trust security, smart home features and services.
The provider edge consists of infrastructure and software owned by the telephone company. Designed by carriers, these “edge clouds” provide infrastructure as a service and 5G enablement within the operator’s network, bringing them closer to the end user. Customers use the telco’s computing and storage resources to automate software that supports the other three edges. Ultimately, these edge clouds will be able to host operations and engagement edge features and data. Typical use cases include edge clouds and private 5G cellular networks.
The edge is tailored for different use cases, but ultimately it only provides computing power and can be used in different scenarios. Hopkins said this would also cause confusion. Part of the code can be run at different edges as needed. For example, Hopkins says some companies are running IoT-type workloads, such as data processing for predictive analytics, on engagement edge computing clusters rather than on the edge servers in the manufacturing floor. This represents the overlap between the engagement edge and the operation edge.
The second duplication occurs when the underlying infrastructure vendor sells computing, storage, networking, and software running at many edges. For example, HPE creates a server that can run on the operational edge, provider edge, or enterprise edge.
And finally, some vendors offer an edge as a service. Because they own the entire compute stack, customers pay only based on the services they purchase.
“Edge as a service is also a marketing term that is realized in different ways by each vendor,” says Hopkins. “Therefore, clients are advised to ignore it and reach the reality of the solution provided. What is the problem it solves? A set of required hardware, network, and software assets. What is?”
---
Autor(en)/Author(s): Daniel Kuhn
Quelle/Source: Illinois News Today, 23.05.2021
Bitte besuchen Sie/Please visit:
