Georgios Kyriakopoulos, VP of Equity Research, SunTrust Robinson Humphrey
Communication networks are inherently complex. The internet itself is a collection of numerous independent networks interconnected to form a network of networks the users of which are able to communicate with each other. For example, when a person on an AT&T network sends an e-mail to someone who uses a different provider such as Verizon, the e-mail must pass from one network to the other to get to its final destination. A data center provides the physical point at which that interconnection can occur. When relatively equal amounts of traffic are exchanged between networks, that exchange is called “peering” and typically network operators set up agreements to trade traffic often at no charge to the other party.
“The closer the CDN server is to the user geographically, the faster the content will be delivered to the user.”
In the early days of the internet, peering took place in public peering points that were established at government and non-profit organizations, known as Network Access Points (NAPs). Over time, many NAPs became part of carrier services and were run by companies such as Verizon, Sprint, and AT&T. With time, the NAPs model became inefficient as it couldn’t scale and keep up with the exponential growth of the internet. Furthermore, the lack of “neutrality” by the carrier owners of these NAPs created a conflict of interest with the participants.
The aforementioned NAPs issues created a market need for network-neutral interconnection points. Today, many customers satisfy their requirement for peering through data center providers because they permit them to peer with the network they require within one location, using simple direct connections. Furthermore, the ubiquitous availability of data centers has aided the proliferation of Content Delivery Networks (CDN), which are systems of distributed servers deployed in multiple data centers via the Internet. CDNs are effective in speeding the delivery of media content, such as video, to users, primarily by taking advantage of their geographic reach. The closer the CDN server is to the user geographically, the faster the content will be delivered to the user.
Additional trends that are influencing the growth of data center and cloud computing include increasing digitization, the proliferation of multiple devices and connections that are driven by the Internet of Things, and the growth of mobility. Furthermore, as the internet consumer population is using more personal cloud storage coupled with growth of machine-to-machine connections, the need for big data analytics results in increased data center traffic. Today, almost two thirds of data center traffic is taking place within data centers on tasks such as moving data from a development environment to a production environment. Given that two thirds of internet traffic is data center related, the importance of data centers both from a colocation and cloud perspective is significant. As such, data center operators that take advantage of scale are more preferable longer-term as they can offer solutions that are scalable, have wide geographic reach, and plenty of connectivity options.
Lastly, the rise of cloud computing and the proliferation of services offered by companies such as Amazon AWS and Microsoft Azure have created the need for more specialized data centers that are often built in-house. Cloud computing allows for the delivery of hosted services over the internet enabling customers to consume computing resources as a utility with flexibility to grow on a real-time basis without having to build and maintain their own infrastructures. While the appeal of cloud services likely increases the construction of in-house mega data centers, the complexity of networks and the need to deliver high performance services with broad geographic reach likely continues to fuel demand for data center services such as colocation that are offered by independent data center operators.