The edge can add another layer of scale and performance and, with the right design, can provide security and help minimize costs. Simply put, it’s where applications and services need presence for digital businesses to be successful — whether they’re media companies, financial institutions, retailers, SAAS providers, or any other kind of enterprise.
Akamai’s platform is a good example — it consists of hundreds of thousands of servers housed in 4,000 locations across more than 1,000 cities in approximately 140 countries. Situated close to billions of users and devices, it’s designed to optimize delivery of web and media content, and run all sorts of web applications. It also provides a security shield with optimized defenses for the services and applications it supports.
How does the edge make a difference?
Capacity at the edge of the internet is growing rapidly. In many cities, broadband connections into homes and offices have tens or hundreds of Mbps (and, in some cities, even Gbps) of capacity, easily supporting high-quality video streams and large software downloads. Internal calculations from Akamai experts added up fixed broadband last-mile connections around the world to yield about 40,000 Tbps of access capacity at the edge. Cellular connections add another 10,000 Tbps or so of capacity in the last mile, and this will grow substantially with 5G. This is a primary reason why the edge is so important–there is a ton of capacity at the edge of the internet.
By comparison, there’s much less capacity in the core of the internet, where big cloud data centers and large backbone providers meet. Using the same approach as above, our internal teams calculated only a few hundred Tbps of capacity for access networks. This imbalance creates a problem when hundreds of millions of users who want to watch video (and do their work!) have access to thousands of Tbps of capacity in the last mile (at the edge) and there’s only a few hundred Tbps of capacity in the core. As more people fire up their favorite online entertainment or interact with other services in clouds, the core gets congested. Lack of capacity in the core makes it hard to service traffic at the edge. Sometimes data centers get congested, but more often it’s the peering points where traffic is exchanged between data centers and end-users that are the choke points.
The high-profile streaming outages that affected some countries last year during the World Cup offer a real-world illustration. As the figure shows, cloud data centers in some regions got swamped and viewers couldn’t see the game as a result. Denial-of-service attacks can cause similar problems, intruding on legitimate user traffic. The good news is delivering traffic from the edge can address these problems because there’s more capacity.
What about 5G?
5G improves the performance of the last mile — with lower latency, higher throughput, and a lot more connected people and THINGS. 5G will also enable lots of cool new apps that will increase internet usage more than ever before. This means there’ll be a lot more traffic and users will expect an even better experience.
Gartner estimates that by 2022, more than half of all enterprise data will be created and processed outside of traditional cloud data centers.1 Today, it’s less than 10%. This begs the question: “What exactly is the edge in 5G and why is it so important?” The answer is that applications will define the edge and without it they’ll be less robust, harder to scale, and costlier.