Our thoughts on the future of digital innovation and the cloud.
Edge to Cloud: How edge computing boosts your productivity
Edge computing. Whether you’ve heard of it or not, you’ve almost certainly used it if you work on a computer. A rapidly growing area of cloud computing, the edge offers intriguing possibilities for fast and effective data provision.
What is the edge?
Edge to the cloud is an essential tool in your digital kit. Typically, it works in concert with cloud or on-premises solutions, that hybrid allowing both pieces to perform complementary functions. While the cloud can handle data volumes and complex analyses, data can only travel so fast. By contrast, the edge allows rapid response to a localized data set by mitigating the distance information has to travel. It’s a sort of just-in-time delivery system for communication.
An everyday instance of the edge is download speed. When visiting a given webpage, your browser likely downloads the images from a content delivery network (CDN), which stores data locally to your network. Images are delivered to your computer more quickly than if they had to travel from a centralized server. From a business point of view, this can translate into real dollars. While the Internet is objectively fast, it’s not always fast enough for the impatience we might feel when a page loads slowly. The CDN – the edge application – mitigates the possibility of losing site visits.
Where does edge computing come from?
Edge computing is older than the Internet. In the mainframe days, 50 years ago, there were relatively few, but expensive, computers. Dumb terminals evolved to access mainframes from remote locations; these allowed for input and output activities, with processing restricted to the mainframe.
Personal computers became more powerful with time, and in the late 1970s, introduced a shift away from mainframe processing. Instead, computing increasingly happened locally, right on the device – an early example of bringing the compute to the edge. Meanwhile, the cloud emerged over the last ten years as a powerful and sophisticated data management tool. Data stores began to centralize in the cloud, much as it was with the old mainframes.
Just as the PC brought processing to the user, so the edge brings the cloud. This kind of local processing allows for unprecedented speed and latency mitigation, ultimately providing more efficient operations and elevating the user experience.
But my processing speed is unprecedented
It’s true, computing is more powerful than we have ever known it. But the edge is not just about being fast, it’s about being fast enough to do unprecedented things with it.
Processing speed may be a simple matter of convenience. Consider the delay time when you ask your voice assistant a question. If Siri or Amazon Echo tapped into a local database rather than transmitting back to the cloud, the gap to response time would be reduced, and the experience smoother and more efficient overall.
Gaps in data transmission can, likewise impact safety. For instance, a self-driving car contains a camera that observes the road and everything on it, tasked with timely responses to those inputs. If the camera sent data to the cloud for processing, it would simply take too long for the car to respond meaningfully. Edge allows the compute to happen locally, and that precious extra nano- or micro-seconds can be life-saving.
Processing is also a matter of efficiency and prioritization. Some operations, such as mine sites, are located where there are no internet or satellite internet results in high latency. This limits the ability to send information to the cloud. However, a “store and forward” approach can mitigate this. By bringing an Azure edge stack to the site, the business can proceed as it would be publishing to the cloud or on-premises. Local data inform on-site decisions, which often need not be retained. Meanwhile, aggregate data is consolidated from multiple sites for future reporting; this smaller data set can then be sent to the cloud when bandwidth becomes available.
Edge and the Internet of Things (IoT)
The edge and IoT go hand-in-hand, as local devices “talk” to local computing centres. Much like store and forward, local processing can act on the data it needs, while key analytical data can be sent back to the cloud.
With IoT, nearly anything can be rigged to send signals to edge devices, bypassing the cloud. Edge stacks, tied together by a common management stream and linked back to a central data location, reduce latency while providing long-term storage and reporting. Integrating these approaches through Azure edge stack, which contains virtual machines running in on-premises racks, not only provides the most efficient outcomes, but also means avoiding the inevitable quarterly changes that third-party solutions requires as equipment evolves.
Does my business need edge computing?
The cloud is powerful but cannot alone solve all data problems. Especially with latency-sensitive applications, it is essential to take a hybrid approach. The edge provides the speed of compute locally. At the same time, the cloud or on-premises solution handles the heavy lifting of analytics and reporting.
The key is that the edge is a latency mitigator. If response times are critical for your operations, it is worth considering as part of your digital solutions.