Neil Ballinger explores the data centre of the future
Data centres are expected to use 20% of the word’s energy by 2025. While companies across the world depend on them to store data, there are new technologies that can improve their energy efficiency and operational effectiveness, if we take advantage of them.
During the dot-com bubble of 1997 to 2000, companies needed fast and consistent internet connectivity to establish a presence online.
Installing the equipment to do this was not viable for many small companies. Internet giants at the time started building large facilities, then known as internet data centres, to provide these businesses with data storage solutions.
Since then, data centres have become smarter and more efficient. But, what emerging technologies are set to shape the next generation of data centres?
Sitting on the edge
According to the DataAge 2025 study, which was recently commissioned by Seagate, almost 20% of data created will be real-time in nature by 2025. This means that manufacturers need to build on their central cloud computing architecture and develop the ability to process and secure more data at the edge.
With edge computing, data analytics is only partly reliant on network bandwidth as most of the computing takes place locally – either in the device itself, the edge data centre or in the fog layer.
Naturally, this will increase the speed at which this data is processed and becomes available.
Huge data centre models won’t become obsolete, but the rise of edge computing could see a large number of small data centres built closer to industrial sites and business parks.
Thermal management of high-power data centres poses a challenge for data centre managers, due to the high operational costs associated with an inefficient facility.
Typically, server rooms in data centres are cooled using classic ambient air-cooling with cold water-recirculation coolers. For high power applications, water-cooled racks are used too.
Water and hybrid air plus water cooled data centres are an alternate cooling solution. This method combines liquid cooling systems, such as rear door heat exchangers located within the racks themselves, with a traditional raised floor cold aisle air cooling system.
Such a solution may be used when the equipment in a data centre is upgraded to higher end or higher power equipment, which may not be manageable with the existing air-cooling system.
Using artificial intelligence
Artificial intelligence (AI) technology has existed for decades but is now going mainstream thanks to the advent of big data, deep learning algorithms and AI-focussed processors.
The technology can now be used in conjunction with data centre infrastructure management (DCIM) software to analyse power consumption, capacity planning and cooling, as well as the overall health of critical systems.
AI can also help to reduce energy consumption. Google recently acquired DeepMind, an AI start-up, to use the technology to reduce costs and improve efficiency in its data centres.
The system was able to achieve a 40% reduction in the amount of energy Google used for cooling the data centre.
According to the International Data Corporation (IDC), half the components in large data centres will offer integrated AI functions by 2022, allowing them to operate autonomously. This, along with the other emerging technologies making their way into data centres, could drastically reduce energy consumption in these facilities, improving their functionality and reliability for businesses across the world.
Neil Bellinger is head of EMEA sales at automation equipment supplier EU Automation.