In 1994, across Europe, a typical website was 10kb in size – little more than a background, a few images and text. Today, a website is typically 1mb in size and public expectation of what it can do has grown into a multi-sensory experience, adding video and audio to basic imagery.
Expectation growth has been fuelled by the ever improving technology that delivers this content to consumers. Ten years ago, most European users were on a dial up connection, receiving internet speeds averaging 700 kbit/s. Today, the global average is 3.1 Mbit/s, and areas of the United States and Europe can receive up to 15 Mbit/s connectivity. In a world with this kind of connectivity, a user can download a website built on video and interactivity in the blink of an eye.
This trend isn’t just refined to the most technologically developed countries; India’s network infrastructure delivers speeds averaging 2 Mbit/s, meaning a demand for lightweight sites is still typical, but internet speeds of up to 15 Mbit/s are becoming available.
With this kind of connectivity, some are beginning to question the need for some long-standing cornerstones of internet architecture. Amongst them is the content delivery network (CDN) – the network of servers distributed throughout a region that allow online content to be delivered faster to end users. The CDN’s role is to shorten the distance between a user and the content they are accessing (a website, for example). The closer the server is, the quicker a user can download that content.
The predicted extinction of this kind of set up makes sense to a point: In a world where last mile internet connectivity is fast enough to deliver large websites without faltering, why create extra architecture to deliver it quicker?
The problem for content providers is that, whilst the last mile of internet delivery is a clear highway for their content to be delivered, the number of people demanding that content is growing exponentially. Every day, more of the world is connected to the internet, the demands of those with internet grow more sophisticated, and households in countries with the heaviest usage add more screens to their connections.
This makes prediction of the CDN’s demise greatly exaggerated. In fact, content delivery networks are becoming more important as more people demand ever more data intensive services online. It’s no longer a case of putting a server close enough for a user to minimise their download time, it’s a case of placing content across a host of different servers to stop thousands of people logging into one overloaded source. A highway, no matter how large, will still be held up by a busy junction at the end, but many smaller roads with well operated junctions will keep traffic flowing.
Netflix is the poster child for the continued need for CDNs. The video streaming service accounts for 34 per cent of peak time traffic in the USA, sending HD video – a large packet of data – to thousands of American homes. Serving so much content to so many homes from a single data centre, or even a few centralised data centres, would results in millions of people watching a series of blocky still pictures instead of the latest Better Call Saul. By distributing content across a range of networked data centres, Netflix spreads that connectivity burden across many servers and eliminates the bottlenecking that it would otherwise risk. Viewers can watch their favourite drug drama in seamless HD.
There will be a time at some point in the distant future when CDN set ups are no longer the most effective delivery method for online content, but the incredible sustained growth of the internet means that time is a long way away, and new business models will still find a need for delivery networks for the foreseeable future. Not to mention the benefits that the CDN’s distributed approach to computing could bring to the Internet of Things, augmented reality and many more innovations that are on the horizon.
What do you think the future of CDNs looks like? Let us know below