Efficiencies improve as computing and storage functions move closer to the data source
Living on the edge is more than a top hit from the early 1990s; it’s how businesses maximize operational efficiency and improve performance for essential applications and cellular IoT connectivity. Edge computing solutions transform how data is handled, processed, stored, and ultimately delivered in environments that demand localized computing power to support IoT and real-time functionality. With multiple edges to choose from, businesses can automate their critical operational functions and decrease downtime through a trusted, scalable computing solution.
The nebulous edge: defining edge computing
The word “edge” may imply a strictly defined border, but edge computing refers to a much more indistinct threshold. The computing edge is a cloud-based network architecture that moves computing functions and storage closer to the data source. Without edge computing, companies must leverage computing resources in remote data centers or the cloud.
To illustrate how the edge works, consider a grocery distribution center. Within that center there are hundreds of applications, including a variety of inventory management systems that each process information through a bank of local servers. To stay online, those servers often require an on-site IT team to troubleshoot and provide ongoing performance updates to keep the servers running. Although there’s a significant cost associated with that upkeep, the servers must be housed locally in order to reduce latency.
Edge computing solutions move those application computing functions from local servers to local “edge” data processing capabilities. As a result, businesses experience improved processing speeds, reduced cost of contextual data storage, reduced cost of wide-area network (WAN) data transmission, and a decrease in costly local servers, all while improving transaction resiliency.
The types of enterprise edge computing
From autonomous vehicles and remote monitoring of gas and oil sensors to predictive maintenance and traffic management, 5G IoT applications are built to run on the edge. However, the edge is not a one-size-fits-all technology. When implementing an edge computing solution, there are three primary options to consider.
The Internet edge
Co-located at Internet points of presence (POPs), the Internet edge takes advantage of cloud-based data centers powered by large server environments such as Equinix. The POPs serve as links for the Internet on route to computing services like Microsoft Azure or Amazon Web Services (AWS) to the cloud.
The Internet edge typically reduces latency to tens of seconds and offers the additional benefits of improved processing speeds for cloud-centric applications, reduced volumes of contextual data storage, and reduced costs of operation.
Remember, because the Internet edge is “outside the firewall,” traffic has to flow through the routing and security stack on the edge router. This induces 1-3 milliseconds of latency right off the bat, before the network transit delay.
The 5G edge
Leveraging towers and remote telecommunication offices, the 5G wireless edge is colocated at cell towers, creating small clouds that are geographically dispersed. With an average latency of only a few seconds, the 5G edge provides the same benefits as the Internet edge, plus it supports end-to-end traffic and lowers costs of WAN data transmission in addition to a reduction of costs associated with operating and maintaining local servers.
Because of its widely distributed connection points, the 5G edge is most useful in environments that are similarly dispersed and require ongoing data collection through sensors. Consider a fast-food restaurant, for example: An average fast-food restaurant has dozens of sensors in each location monitoring grease levels, temperatures, refrigeration, cleanliness, and more. These sensors create a significant amount of chatter that doesn’t necessarily need to be shipped across the WAN. By employing 5G edge solutions close to their locations, a fast-food chain can reduce backhaul and become more efficient while saving money and improving their quality of food and service.
As with the Internet edge, the 5G edge is outside the customer premise firewall and subject to the same 1-3 milliseconds of latency incurred by just traversing the edge router.
The customer edge
The customer edge is an on-premises edge that utilizes specially equipped routers to run dynamic applications in virtual machines or containers. Because this edge is closest to the data source, latency is reduced to mere milliseconds and customers receive all the benefits of the Internet and 5G edges in addition to improved transaction resiliency.
Use cases that require immediate feedback — such as video surveillance and patient monitoring — are best suited for the customer edge. Older connected devices, like the handheld Telxon Terminals, that transmit a significant amount of data in each transaction using the American Standard Code for Information Interchange (ASCII) will also be the most efficient on this edge.
An added benefit of the customer edge is that virtual edge computing platforms such as Cradlepoint’s NetCloud Service and cellular wireless routers are available in models capable of using containers to run edge applications. These routers are easy to set up, simple to monitor, and highly secure without the need for a separate server. In addition to containers, edge computing can be delivered through custom Python apps or connectors such as Microsoft Azure or AWS.
One of the real benefits of the customer premise model is that the edge computing happens inside the firewall, eliminating the 1-3 millisecond router transit tax.
Final considerations when choosing the best edge for your business
Depending on business requirements and objectives, an enterprise edge computing portfolio may include multiple solutions. Regardless of which edge is selected, it’s important to consider that each solution will have to go through security and Quality of Service (QoS) stacks in the router, which will ultimately add latency. That being said, 5G connectivity provides fiber-fast speed and ultra-low latency, allowing more applications to run at the 5G edge than ever before.