As we head into a new decade what will the future hold for data centres? With the 2020s set to be the ‘decade of data’, Dave Sterlace, global head of technology, data centres at ABB, outlines his top five trends to look out for in 2020 and beyond.
The 2010s saw a huge increase in digitalisation, and with most of the world’s IP going through data centres, demand for more digitally-enabled, responsive facilities has also grown.
In 2020, it is likely we will see a number of data-hungry technologies becoming more commonly used. With this in mind, what will be the top five drivers for the data centre sector next year?
Sustainability concerns will drive investment decisions
As some of the largest consumers of energy, the focus for 2020 and into the next decade will be on how data centres can reduce their consumption and carbon footprint – particularly against a backdrop of increased climate awareness.
As more operations adopt IoT platforms – and 5G enabled networks become a reality – data demand will grow significantly, and with that will come a major increase in energy consumption. The industry will shift to how providers can operate in the most sustainable way, without compromising on operational reliability and efficiency. A holistic approach to data centre design will look to where energy efficiencies can be gained through small improvements in electrification and digitalisation.
With sustainability likely to remain a key economic and political driver in the years ahead, data centres can adapt with digital solutions to save cost and carbon, without compromising critical business operations.
2020 will be the year 5G starts to have a real impact
Without a doubt, 5G will be one of the biggest disruptors for the data centre industry. Although it is still in its infancy, the global rollout of 5G will put the growth of digital data into overdrive, with Cisco predicting that we will soon enter the ‘mobile zettabyte era’.
Everything will be impacted by 5G, from high-performance cloud data centres to edge services. With the capability of supporting one million devices per square kilometer, it really will bring forward the IoT age - and with that will come a ‘data tsunami’ that will need to be collected, analysed and stored very quickly.
As a result, data centres will need to adapt to the demands 5G will present. It will require them to improve and upgrade processes and infrastructure to manage the speed of content. As such, we’ll see the increase of the ‘smart’ data centre and the greater use of the cloud, and other technologies such as edge computing, to manage demand.
The continuing rise of the edge data centre market will drive greater autonomy
The IoT age is driving the need for data to compute at the edge - according to Pricewaterhouse Cooper, the global market for edge data centres is expected to nearly triple to $13.5 billion in 2024 from $4 billion in 2017.
5G is also expected to enable new data-intensive services like autonomous vehicles. Some estimates predict these will generate terabytes of data per day, which will need to be processed very quickly. Any lack of connectivity – even for a short period of time – could have significant impact. Therefore, the edge data centre will be a necessary enabler for autonomous vehicles and compatible technologies that rely on data such as virtual reality and augmented reality, and the infrastructure will be critical in moving these types of technologies from novel gadgets to critical business assets.
However, the reality is that every edge data centre cannot have staff on site, therefore, tools such as digitalisation will support these facilities by providing peer to peer communications and offering greater visibility and granular data to help remote managers make responsible and speedy decisions to diagnose issues.
Managing safety in response to demand for more data centre capacity will be a major concern, leading to increased use of prefabricated solutions
Demand for data means there will be a need for data centres that can be both constructed and fully operational in increasingly tighter timescales.
To help mitigate the risks posed by constructing data centres as quickly as possible, it is likely we see an increase in the use of offsite build. With a complicated data centre project that has short delivery timelines, pre-engineered product packages, along with prefabricated solutions, may be the answer.
With these solutions, a data centre operator can benefit from prefabricated and pre-tested modules for power distribution solutions that reduce risk exposure during site works. These could include modular systems solutions, such as prefabricated eHouses and skid-mounted unit substations, to deliver switchgear, transformers, ancillaries and other electrification components.
By offering great flexibility, a high level of safety and integration of intelligent technology, these solutions ensure that a data centre is reliable and always available, and reduce the time and logistical factors needed to construct a data centre.
AI, AR and machine-learning will be used to optimise data centre efficiency… gradually
Intelligent technologies will continue to have a major impact on the data centre sector. For example, AI can have a real impact on predictive maintenance, machine-learning can help refine predictive maintenance by learning from exceptions, and AR can couple with both to make ‘remote hands’ for maintenance a reality.
However, giving a non-human, machine-learning algorithm control of mission-critical infrastructure is not something that can happen overnight. For many operators, relinquishing control and putting faith in an AI-based system is a step they’re not quite ready to take. That said, the potential benefits are considerable, particularly in terms of increasing efficiency and freeing up engineers to focus on other crucial tasks such as maintenance and safety.
What is important to consider is – as with human skills – machine-learning technology needs to be ‘trained’ for the job it is there to carry out, so it can effectively respond to the specific functions, needs and pressures at each individual data centre.
So, while these technologies can have a real, positive impact on the efficiency of a data centre, the use of AI-based automated control systems will be gradual.