What is Edge Computing?

The practice of placing, processing power near the “edge” of a network, so that data can be processed locally, before entering a Wide Area Network.


What it is: In Edge Computing, processing power is placed near the “edge” of a network, so that data can be processed locally, rather than needing to be passed across a Wide Area Network. This is different from typical cloud computing, in which data is routed to a centralized location for processing. By localizing data, Edge Computing can reduce data transfers, increase response times and potentially increase security and resilience.

What it does: Edge Computing places computing power close to where it is needed, whether by locating compute in actual IoT devices or by placing it in server devices closer to the edge. This speeds up processing, which is vital in some use cases—such as autonomous vehicles—and which can improve the consumer experience. Moreover, it avoids data traffic jams that can occur in Wide Area Networks, as well as potential outages in centralized computing locations.

Why it matters: In the era of the cloud, it’s easier than ever for companies to scale their computing power. However, as this continues, computing power is concentrated, which places stress on networking capacity, creating architectural issues. Meanwhile, smart buildings, smart vehicles, and many other innovations of the IoT era rely on fast, reliable bi-directional transmission of information. Edge Computing enables this.

What to do about it: If your products, whether applications or devices, collect and process high quantities of data near the edge, consider Edge Computing solutions as a way of lowering latency. Note that services such as Amazon CloudFront make it easy to integrate edge servers with centralized compute. Keep in mind that Edge Computing can cause security risks since decentralizing computing power creates a greater attack surface.

Business Advantages

  • Decreases network latency, reduces bottlenecks and increases efficiency by decentralizing computing power
  • Enables simple or complex processing of data to take place locally, from deduplication to application of machine learning
  • Provides the ability to cache in the case of network outage.
  • Increases the viability of data-intensive IoT appliances, like smart cars and smart homes
  • Can lower costs by reducing the amount of data processed by centralized compute services

Use Cases

  • Can make smart homes and spaces better by decreasing latency, making automatic systems respond to inhabitants in milliseconds rather than seconds
  • Can increase the efficiency and speed of automated manufacturing, by producing faster responses to changes in local conditions
  • Can improve Augmented Reality experiences by speeding up computationally intensive processes that depend on locally translated data
  • Can improve speed of video processing and other high-bandwidth activities

Edge Computing and Security

By its very nature, Edge Computing creates more potential for attack by locating networked computing power in more devices, some of which might be physically accessible by malicious actors and which require to be secured. In addition, the new use cases may create unplanned risks, for example stealing data locally which would not have existed before. These challenges are not insurmountable, but they require assessment and vigilance, or initially inexpensive Edge Computing solutions may increase the chances of incurring devastating long-term security costs. However, Edge Computing also offers potential security benefits, since decentralized computing power could make attacks on centralized data stores less likely or impactful.

Case Study

Recently, HPE partnered with Texmark Chemicals to design smarter refineries. This involved combining sensors on mechanical assets with Edge Computing, enabling high-speed analysis of locally collected data. Access to this analysis improved productivity and worker safety by creating a more accurate picture of plant functioning.