What is cloud computing?
Cloud computing places your workloads and apps onto a vendor’s servers in a centralized data center, making them globally accessible.
Cloud computing emerged as enterprises moved their workloads and apps outside of centralized on-premises data centers and onto cloud servers and hardware rented from cloud platform providers. The cloud has continued to grow in popularity, as organizations appreciate the computing flexibility of near-limitless scaling.
The resources that organizations rent from cloud platform providers can take a number of forms, such as:
- Virtual instances of hardware (historically, the most common)
- Pure compute time to run serverless or other workloads
- Data storage
Most cloud service providers — such as Akamai, Microsoft Azure, Google Cloud, and Amazon Web Services — offer all these options and more, providing a full feature set upon which whole organizations can be built.
What are the benefits of cloud computing?
The key benefits of cloud computing include:
- Near-limitless scaling: Cloud computing has allowed organizations to scale their computing resources on demand, without large up-front hardware costs. Cloud computing makes massive amounts of computing power available as needed.
- Flexible pricing: Besides scaling up, cloud computing allows organizations to scale down flexibly — saving costs and making organizations more agile.
- Simplified operation: You no longer need to manage the hardware, the operating system, or even the virtual machine; this can all be delegated to the platform provider, freeing your organization to focus on its business.
- Simplified management for distributed users: Rather than managing multiple on-premises centers for each office in a different region, it’s possible to share cloud infrastructure across users anywhere there is an internet connection.
With cloud computing, adding new IT infrastructure can be as easy as the press of a button.
Examples of cloud computing providers
Akamai Connected Cloud and AWS are examples of cloud computing platforms that provide a wide range of compute, storage, and database options. These cloud computing services support networking and security features that organizations find are more cost and operationally efficient to outsource.
What is edge computing?
Edge computing is a method of processing data close to users and devices. Workloads are distributed and executed as close as possible to the request.
By locating workloads as close as possible to the end user, the edge computing approach saves bandwidth costs and reduces latency, resulting in the high-speed, economically scalable digital experiences that people have come to expect.
Edge computing is still evolving, and the actual locations vary.
Examples of edge computing infrastructure include a dedicated edge server, a network of edge servers, and even an Internet of Things (IoT) device (Figure 1).
Content delivery networks (CDNs) are also seen as a type of edge network; in fact, they’re the precursor to distributing compute. CDNs serve web and video content, media, images, APIs, and more to users more quickly by caching it closer to them. In many cases, traditional CDNs evolved into multifunctional edge networks with edge servers that can also run edge computing workloads.
Location matters
The key concept in edge computing is that location matters. An edge workload cannot live in a data center in a network location that’s far, far away from the end user. Instead, locating workloads closer to the end user, and at the edge of the network, can help improve the user experience while also personalizing it.
What are the benefits of edge computing?
The key benefits of edge computing include:
- Reduced latency: Just like the ATMs of a large bank with a global footprint, you can always find an edge server close by. This means that network requests and operations take less time, leading to faster response times.
- Interactivity: With reduced latency, digital experiences become more immersive and responsive.
- Personalization: Each edge server can contain data and processing specific to its location, offering user experiences that can be localized and customized.
- Cost savings and performance: Handling data closer to its source can reduce cloud bandwidth requirements.
Examples of edge computing
An example of edge computing is distributed processing for IoT devices. By processing the data for IoT devices closer to the source, less cloud bandwidth is used, and only the relevant data needs to be sent onward to the main database. Additionally, with reduced latency, battery-constrained IoT devices can conserve energy by reducing total transmission time.
A small vacation rental property may have a few dozen IoT devices while a small connected city may have 100,000-plus devices. Those devices likely need to update their status constantly. Although we could scale one’s centralized cloud infrastructure to handle all those requests, we would run into the same prohibitive cloud cost economics that we saw when CDNs were invented. With edge computing, you can handle IoT volumes more cost-effectively without sacrificing functionality, performance, or availability.
Another example is in caching localized data. By caching data relevant to users in a specific location or region at the edge, latency is reduced, and the experience can be personalized.
In addition, edge computing is changing how enterprises analyze real-time data to mine customer insights. For all these reasons, it’s no surprise that spending on edge computing continues to increase.
Edge computing vs cloud computing: Similarities and differences
Cloud and edge computing share some key similarities:
- Enable applications and business workloads: Both cloud and edge computing can execute business logic and support running workloads and applications.
- Streamline analytics: Both cloud and edge computing can be used to receive, ingest, and process data.
- Usage-based pricing models: Both cloud and edge computing feature pricing models which typically scale with usage. This is in contrast to owning and operating your own infrastructure, which typically has higher initial capital costs.
Edge computing and the public cloud
While edge computing and cloud computing are often adopted together strategically (as we’ll see below), understanding what makes each one distinct is important so that each paradigm can be applied appropriately. So then, how is edge computing separate from the public cloud?
- Time sensitivity: Edge computing, being physically closer, can respond faster than cloud computing can.
- Data volume: Edge computing can process data closer to the source before interacting with the cloud, which saves cloud bandwidth and lessens compute requirements.
Based on this explanation, it might seem like edge computing is always more advantageous than cloud computing. That brings us to a commonly asked question about the two.
Does edge computing replace cloud computing?
Edge computing and cloud computing are coexisting technologies (not competing); neither one is better than the other. In fact, many use cases are best served by combining the two.
For example, autonomous vehicles, which generate a massive amount of data, may use edge computing for close-proximity data processing and decision-making while data that is useful for long-term analysis and machine learning (ML) model training may be pushed to the cloud. In healthcare, edge computing with artificial intelligence (AI) is being used to support real-time patient monitoring and control of IoT medical devices. Meanwhile, aggregated data is sent to the cloud for analysis and research.
Because edge computing involves a diverse array of devices, components, and platforms, open software and standards promote interoperability by providing a common language and protocols. Open software and standards provide developers with flexibility to build and customize, and they help prevent vendor lock-in by promoting innovation and competition so that businesses have more freedom to choose among different providers and technologies.
What is fog computing?
While edge computing and cloud computing are well-known and oft-adopted technologies, a related approach has emerged in recent years that is worth exploring: fog computing.
Fog computing is a computing infrastructure model that seeks to bridge the gap between cloud computing and edge computing. It distributes data, compute, storage, and applications in an efficient manner between the data source and the cloud. That distribution among resources includes the use of edge devices, regional cloud servers, and traditional cloud data centers.
Fog computing enables the quick response times of edge computing, along with a reduction in the amount of data that needs to be sent to the cloud for processing or storage. Centralized data storage and computing in the cloud is still available and used when needed. Fog computing is still in its early stages and may be subsumed eventually by programmable edge.
When to use cloud computing and when to use edge computing
Choosing between edge and cloud computing depends on the specific requirements and use cases. Cloud computing is all about ease of access and near-limitless scale — one of its most popular use cases is for IoT.
When to use cloud computing
Here are some of the situations in which you can benefit from the advantages of cloud computing:
- High workload size: Cloud computing provides the ability to handle a variety of workload sizes, including those that are not feasible on edge infrastructure.
- Data analysis: Cloud computing can provide powerful data analysis and ML capabilities, enabling organizations to gain insights from their massive datasets and make better business decisions.
When to use edge computing
Edge computing is about removing distance from the server to the user. Here are some of the situations in which you can benefit from the advantages of edge computing:
- Requirement for low latency: Applications that require real-time processing power — such as autonomous vehicles, robotics, or industrial automation — need to minimize latency. Edge computing is a good fit for these applications, as data can be processed locally without having to transmit it to a faraway cloud data center.
- Limited bandwidth: In scenarios with limited bandwidth, such as in remote locations or areas with poor internet connectivity, edge computing can be a better option as it reduces the amount of data that needs to be transmitted to the cloud.
- Security and privacy: Applications that require high levels of security and privacy, such as healthcare or financial services, can benefit from edge computing. By processing and storing sensitive data closer to the source, such data is less vulnerable to cyberattacks or data breaches.
- Cost: In scenarios with high bandwidth or cloud storage costs, edge computing can be a more cost-effective option. Edge computing reduces the amount of data that needs to be transmitted to the cloud, thereby reducing bandwidth requirements and lowering costs.
A combined approach
In many scenarios, a may be the best choice. You can continue to use cloud computing for processing and main data storage, but add edge computing for additional capabilities and refinement. In this approach, neither replaces the other; instead, you’ll use cloud computing and edge computing together for advanced use cases and potential cost savings.