The Death of the Data Center: Why We Were All Wrong

Not so long ago, it looked like many businesses were preparing to say goodbye to one of the staples of their IT department - the data center. With more firms looking towards cloud computing for most or even all of their needs, some in the industry began questioning whether the days of the traditional on-premise data center were numbered.

Among those forecasting the death of the data center was tech industry analyst Gartner. The company predicted back in 2019 that by 2025, four out of five enterprises would have shuttered their traditional data centers. It cited issues such as the cost of building and maintaining these facilities as a key driver for many, as well as the fact that many mission-critical operations are now able to be accessed easily via the cloud.

But was this obituary premature? In fact, many commentators are suggesting that, far from disappearing, the data center is still likely to have a place in many enterprises. It may look a little different than in the past, but it's increasingly clear that the cloud-only future isn't for everyone.

Why cloud-only may not always be the answer

On the surface, cloud computing is still an attractive option for many firms. As previous concerns about security and reliability have faded away in the face of growing evidence that these issues aren’t a major problem, demand for cloud has been growing hugely in recent years. For instance, IDC forecasts global public cloud revenue is set to rise from $180 billion in 2018 to $277 billion by the end of this year.

But this may not always offer the most convenient or cost-effective solutions. For many businesses, cloud services don't always offer the huge savings they expect. While a shift from traditional CAPEX to OPEX expenditure may be useful for firms where cash flow is tight, long-term service costs can mount up and, depending on your needs, may end up costing you more.

Meanwhile, some firms may have a need for a data center as part of their backup and recovery strategy, or to make it easier to keep compliance with industry-specific regulations. What's more, migrating to the cloud is a complex and time-consuming process. If you have to significantly rearchitect applications or data sets to maintain functionality in the cloud, you have to consider whether this effort will be worth it. If not, on-premise will still be the way to go.

Could colocation offer a better alternative for firms?

One increasingly popular solution for many firms is colocation, where the business rents space within a data center managed by a third party. This may sound similar to cloud computing, but the key difference is the business still keeps ownership of their hardware and other equipment, while the colocation partner provides rack space, cooling, power, bandwidth and physical security.

This can offer many advantages over a traditional on-premise data center. Firstly, it means businesses can pay for exactly what they need, rather than having to plan their own data centers that run the risk of being too big or small for their current needs. Next, it reduces expenditure on issues such as power and cooling. As the costs of these are spread over multiple colocation users, the economies of scale deliver benefits.

It may also be useful from a sustainability perspective, which is a factor that's increasingly important to many businesses. Tackling the environmental impact of data centers is now a key part of green IT, and by sharing resources wherever possible, these effects can be minimized.

The key trends driving tomorrow's data center

With data centers set to be with us for the foreseeable future, it's therefore important to keep up to speed with the latest developments and trends in order to make the most of these resources.

Hyperscaling

Hyperscale data centers will be essential for large enterprises as the amount of data they deal with continues to grow exponentially. An effective data center will therefore need to be able to    respond rapidly to changing demands and easily scale up without replacing entire servers.

Edge computing

One of the growing challenges for many businesses is data that isn't stored or processed in a centralized data center. Edge computing calls for processing to take place as close as possible to its point of use - for example within Internet of Things devices. This means that instead of routing traffic to and from a single large data center, businesses will have to develop smaller, more agile solutions that can sit throughout the business to handle local traffic.

AI and automation

The use of artificial intelligence (AI), machine learning (ML) and automation will also all be essential to the efficient running of future data centers. Automation tools will prove especially useful where activities such as monitoring and optimizing need to take place remotely, and will also be necessary to cope with expected skills shortages when it comes to data center operation. Meanwhile, AI and ML will have key roles to play in helping spot patterns and optimize operations, making them faster and more efficient.