Edge computing refers to computing that takes place at or near the user’s or data source’s physical location. Users benefit from faster, more reliable services, and businesses benefit from the flexibility of hybrid cloud computing by bringing computing services closer to these locations. Edge computing is one method for a company to use and distribute a shared pool of resources across multiple locations.
Role of Edge Computing in Datacenters and the Public Cloud?
Edge computing is a strategy for extending a consistent environment from the core datacenter to physical locations close to users and data. An edge strategy extends a cloud environment out to many more locations, similar to how a hybrid cloud strategy allows organizations to run the same workloads in their own datacenters and on public cloud infrastructure (such as Computeman).
Many industries, including telecommunications, manufacturing, transportation, utilities, and others, are using edge computing today. People use edge computing for a variety of reasons, as varied as the organizations they support.
Some of the most common edge use cases
Many edge use cases stem from the need to process data locally in real time—situations in which sending data to a datacenter for processing would result in unacceptable latency.
Consider a modern manufacturing plant as an example of edge computing driven by the need for real-time data processing. Internet of Things (IoT) sensors on the factory floor produce a steady stream of data that can be used to prevent breakdowns and improve operations. A modern plant with 2,000 pieces of equipment can generate 2,200 terabytes of data per month, according to one estimate. Processing that trove of data close to the equipment is faster—and less expensive—than sending it to a remote datacenter first. However, connecting the equipment through a centralized data platform is still desirable. Equipment, for example, can receive standardized software updates and share filtered data that can aid in the improvement of operations in other factory locations.
Another common example of edge computing is connected vehicles. Computers are installed on buses and trains to track passenger flow and service delivery. With the technology onboard their trucks, delivery drivers can find the most efficient routes. When using an edge computing strategy, each vehicle runs on the same standardized platform as the rest of the fleet, improving service reliability and ensuring data security across the board.
Another example of edge computing is autonomous vehicles, which process a large amount of real-time data in a situation where connectivity may be intermittent. Due to the large amount of data, autonomous vehicles, such as self-driving cars, process sensor data on board to reduce latency. They can still connect to a central location for software updates over the air.
Edge computing also aids in the smooth operation of popular internet services. Content delivery networks (CDNs) place data servers close to users’ locations, allowing busy websites to load quickly and enabling fast video streaming.
A nearby 5G cell tower is another example of edge computing in action. Network functions virtualization (NFV), which uses virtual machines running on standard hardware at the network edge, is becoming increasingly popular among telecom providers. These virtual machines can take the place of costly proprietary hardware. With an edge computing strategy, providers can keep software running consistently and with uniform security standards in tens of thousands of remote locations. In a mobile network, applications running close to the end user reduce latency and allow providers to offer new services.
Benefits of Edge Computing
Edge computing can result in cheaper, faster, and more stable services. Edge computing provides a faster, more consistent experience for users. Edge means low-latency, highly available apps with real-time monitoring for enterprises and service providers.
Edge computing can lower network costs, avoid bandwidth constraints, shorten transmission times, reduce service failures, and give you more control over sensitive data movement. Load times are reduced, and online services are brought closer to users, allowing for both dynamic and static caching.
Computing at the edge benefits applications that benefit from a faster response time, such as augmented reality and virtual reality.
The ability to conduct on-site big data analytics and aggregation, which allows for near real-time decision making, is another advantage of edge computing. By keeping all of that computing power local, edge computing further reduces the risk of sensitive data being exposed, allowing businesses to enforce security practices and comply with regulatory policies.
The resiliency and cost savings associated with edge computing benefit enterprise customers. Regional sites can continue to operate independently from a core site by keeping computing power local, even if the core site goes down for some reason. By keeping compute processing power closer to its source, the cost of paying for bandwidth to transport data between core and regional sites is greatly reduced.
An edge platform can help with operations and app development consistency. In contrast to a datacenter, it should support interoperability to account for a greater mix of hardware and software environments. In an open ecosystem, an effective edge strategy also allows products from multiple vendors to work together.
Parts of an Edge Network
Edge computing can be visualized as a series of circles radiating outward from the code data center. Each one represents a tier that is getting closer to the edge.
Provider/Enterprise Core
These are the conventional “non-edge” tiers, which are owned and operated by public cloud providers, telecommunications service providers, or large enterprises.
Service Provider Edge
These tiers are located between core or regional datacenters and last mile access points, which are typically owned and operated by a telco or internet service provider and from which the provider serves multiple customers.
End-user Premises Edge
On the end-user side of last mile access, the edge tiers can be enterprise (e.g., a retail store, a factory, or a train) or consumer (e.g., a residential household, a car).
Device Edge
Individual (non-clustered) systems that connect sensors/actuators directly using non-internet protocols. This is the network’s outermost node.
Edge computing, AI/ML, and Data analytics
With its emphasis on data collection and real-time computation, edge computing can help data-intensive intelligent applications succeed. For instance, artificial intelligence/machine learning (AI/ML) tasks such as image recognition algorithms can be performed more efficiently closer to the data source, eliminating the need to shuttle large amounts of data to a centralized datacenter.
These applications take numerous data points and infer higher-value information that can assist organizations in making more informed decisions. This functionality can be used to enhance a variety of business interactions, including customer experiences, proactive maintenance, fraud prevention, and clinical decision making.
By treating each incoming data point as an event, organizations can use decision management and machine learning/artificial intelligence inference techniques to filter, process, qualify, and combine events in order to derive higher-order information.
Data-intensive applications can be divided into a series of stages, each of which is performed at a different location within the IT landscape. The data ingestion stage—when data is gathered, pre-processed, and transported—is where the edge comes into play. The data is then processed through engineering and analytics stages — typically in a public or private cloud environment — where it is stored and transformed before being used to train machine learning models. Then it’s back to the edge for the runtime inference stage, which serves and monitors those machine learning models.
A flexible, adaptable, and elastic infrastructure and application development platform is required to meet these diverse needs and to connect these various stages.
A hybrid cloud approach, which ensures a consistent user experience across public and private clouds, enables optimal provisioning of data collection and intelligent inference workloads at the edge of an environment, resource-intensive data processing and training workloads across cloud environments, and business events and insight management systems close to business users.
Edge computing is a critical component of the hybrid cloud vision, as it enables a unified application and operation experience.
Telecommunications
Many telecommunications service providers are prioritizing edge computing as they move workloads and services closer to the network’s edge.
When it comes to serving high-demand network applications such as voice and video calls, milliseconds matter. Due to the fact that edge computing can significantly reduce the effect of latency on applications, service providers can offer new apps and services that enhance the experience of existing apps, particularly in light of 5G advancements.
However, it is not simply about adding new services. Providers are increasingly relying on edge strategies to streamline network operations and boost flexibility, availability, efficiency, resiliency, and scalability.
What is network function virtualization (NFV)?
NFV is a strategy that applies IT virtualization to the use case of network functions. NFV enables the use of commodity servers to perform functions that previously required expensive proprietary hardware.
What exactly is vRAN?
Radio access networks (RANs) are nodes in an operator’s network that connect end-user devices to the rest of the network. Similarly to network functions, RANs can be virtualized, resulting in the virtual radio access network, or vRAN.
The ongoing deployment of 5G networks frequently relies on vRAN to streamline operations, serve more devices, and support more demanding applications.
What exactly is MEC?
MEC stands for multi-access edge computing, a technology that enables service providers to offer customers an application service environment at the mobile network’s edge, close to users’ mobile devices.
MEC has several advantages, including increased throughput and decreased latency. MEC provides connection points for application developers and content providers, as well as access to lower-level network functions and information processing.
Edge Computing Relation to Cloud Computing
Cloud computing is the process of executing workloads within clouds—information technology environments that abstract, pool, and share scalable resources across a network.
Historically, cloud computing has been centered on centralizing cloud services in a few large datacenters. Centralization enabled highly scalable and efficient resource sharing while maintaining control and enterprise security.
Edge computing addresses use cases that are insufficiently addressed by cloud computing’s centralization approach, frequently due to networking requirements or other constraints.
Additionally, a cloud strategy based on containerized software complements the model of edge computing. Containers enable businesses to run apps wherever it makes the most sense. A containerization strategy enables an organization to move applications between the datacenter and the edge, or vice versa, with minimal operational impact.
What is IoT, and What is Edge Devices?
The Internet of Things (IoT) is a term that refers to the process of connecting everyday physical objects to the internet. These objects range from common household items such as lightbulbs to healthcare assets such as medical devices, to wearables, smart devices, and even smart cities.
Not all IoT devices are edge devices. However, many organizations use connected devices as part of their edge strategies. Edge computing can be used to add compute power to the edges of an IoT-enabled network in order to reduce communication latency between IoT-enabled devices and the central IT networks to which they are connected.
The simple act of transmitting or receiving data ushered in the Internet of Things. However, using edge computing to send, receive, and analyze data in conjunction with IoT applications is a more modern approach.
How about the IIoT?
Industrial Internet of Things (IIoT) is a related concept that refers to industrial equipment that is connected to the internet, such as machinery that is used in a manufacturing plant, agriculture facility, or supply chain.
Fog Computing and its Relation to Edge Computing?
Fog computing is a term that refers to computing that occurs in distributed physical locations, near users and data sources.
The term “fog computing” is a colloquial term for “edge computing.” Apart from terminology, fog computing and edge computing are synonymous.
Challenges of Edge Computing?
While edge computing can simplify a distributed IT environment, implementing and managing edge infrastructure is not always straightforward.
- Scaling out edge servers to numerous small locations can be more challenging than adding equivalent capacity to a single core datacenter. Increased overhead associated with physical locations can be challenging for smaller businesses to manage.
- Typically, edge computing sites are located in remote locations with little or no on-site technical expertise. If something goes wrong on-site, you need an infrastructure in place that can be easily repaired by non-technical local labor and then managed centrally by a small number of remote experts.
- To simplify management and facilitate troubleshooting, site management operations must be highly reproducible across all edge computing sites. When software is implemented slightly differently at each location, complications arise.
Physical security at edge locations is frequently significantly less than at core locations. A strategy on the cutting edge must account for a higher risk of malicious or accidental situations.
As data sources and storage become more dispersed, organizations require a unified horizontal infrastructure that spans their entire IT infrastructure, including edge sites. Even for organizations accustomed to operating across multiple geographies, edge computing introduces new infrastructure challenges. Businesses require edge computing solutions that:
- Can be managed in the same way as their centralized infrastructure using the same tools and processes. This includes the automated provisioning, management, and orchestration of hundreds, if not thousands, of sites with little (or no) IT staff.
- Address the requirements of various edge tiers, each of which has a unique set of requirements, including the size of the hardware footprint, the difficulty of the environment, and the cost.
- Allow for the use of hybrid workloads composed of virtual machines, containers, and bare-metal nodes for network functions, video streaming, gaming, artificial intelligence/machine learning, and mission-critical applications.
- Ascertain that edge sites remain operational in the event of network failures.
- Are compatible with components from a variety of vendors. No single vendor is capable of providing a complete solution.
How can Computeman help with edge computing?
Computeman’s broad portfolio enables platform, application, and developer services through connectivity, integration, and infrastructure. These robust building blocks enable customers to address even the most difficult use cases.