By: Michael Rabinowitz
We live in a hyper-connected world. Today, there are billions of connected devices, a trillion sensors in everyday objects, and more mobile subscriptions than people on the planet.
As the Internet of Things (IoT) gathers pace and mobile operators begin rolling out lightning-fast 5G networks, pressure will only continue to grow on the connectivity backbone supporting the modern world. Global data is now doubling every two years and the volume of information created since 2013 outweighs everything else produced in the entire history of mankind.
As the nexus for all the connectivity, processing and storage underpinning business in the digital age, the data centre lies at the heart of meeting organisations’ needs.
The leading edge
Our increasingly connected world is driving the rise of ‘edge computing’ – that is, the need to move data processing to the edges of the network, closer to both end-users and devices. This move is essential because much of the IoT is extremely latency-sensitive.
As Ian Hughes, an IoT analyst at 451 Research, comments: “Edge computing is an architectural necessity to deal with the sheer volume of IoT data and the need to act upon that in a timely fashion”. Autonomous vehicles are a good example here. Driverless cars must make extremely rapid decisions in life-and-death situations. There’s no time to process information remotely in distant cloud servers; computation needs to happen close by, enabling responses in tens - not hundreds - of milliseconds.
Today, connectivity providers need to consider how they will meet the demands of the IoT and edge computing. For instance, 5G mobile networks are now being designed with the IoT’s low-latency and rapid response times in mind. Multi-layered networks and new radio access technology will enable 5G to support high throughputs of up to 10gb/s at a latency as low as 1ms. However, connectivity providers also need to consider the physical location of their data centres as part of this puzzle, ensuring rapid access to powerful processing on the network edge.
With economies of scale causing many IoT use-cases to roll out in dense urban locations first, data centres will be needed in the heart of the world’s busiest cities to enable successful edge computing strategies.
While high real-estate prices make it costly for organisations to build data centres in urban centres, colocation offers a better alternative. Carrier and cloud neutral facilities provide highly connected, secure, scalable infrastructure with minimal overheads and maximum convenience. These facilities also make it possible to roll out a multitude of ‘micro-data centres’ in different locations, enabling connectivity providers to support data processing at the network edge far more cost-effectively.
In contrast, high-latency makes regional data centres ill-suited to the demands of edge computing. A remote data centre location can also mean inconvenience and delays when it comes to installing, upgrading or repairing critical systems, while a centrally located urban facility is both highly accessible and able to ensure 24/7 access to expertise.
Colocate in the capital
As the need for edge computing increases, Interxion is seeing ever more customers choose to colocate in central London. To support this growing demand, we’re expanding our central London campus with a £30 million investment in a third data centre. LON3 will add 1,800 sqm to our uniquely located site, strategically positioned between the Square Mile and Tech City.
As a global hub for telecoms companies, cloud providers and international fibre optic links, central London is the perfect place to pilot IoT projects and develop new opportunities. The critical mass of people and devices nearby means connectivity providers are ideally placed to take advantage of emerging technologies and services to solve challenges, enhance competitiveness and thrill customers.
See how Interxion’s expanding London data centre campus can help your organisation gain the network edge, with processing and connectivity in the heart of the capital.