The data centre market has undergone considerable growth in recent years with digital transformation rampant among both personal and working lives.
However, one impact of this development is the need for more data centres, resulting in an enormous increase in the amount of energy being consumed to the point where it has become a global concern.
In light of this, we spoke with SPIE UK data centre director Peter Westwood to garner his thoughts on trends and technologies that are emerging that data centres can employ to improve efficiency.
The first trend that was brought up was edge computing.
“The evolution of edge computing, and decentralising with the extension of campus networks, cellular networks, data centre networks, or the cloud, and a more decentralised approach is required to address digital business infrastructure requirements. The idea being that as the volume and velocity of data increases, so too does the inefficiency of streaming all this information to a cloud or data centre for processing,” says Westwood.
“Data centre companies are working hard to decentralise compute power and place it closer to the point where data is generated. This process has started to lead the requirement to micro-data centres, branch locations, and smaller hubs to process data. The future designed edge ecosystem architectures will lead to greater efficiency and resiliency.
Westwood’s next trend is the convergence of systems.
“This brings the four core aspects of a data centre compute, storage, networking and server virtualisation into a single chassis. Hyper-converged infrastructure (HCI) adds tighter integration between more components through software. Converged Technologies has been a trend, and with these types of architectures it aims to remove silos of resources, challenges around administration, and issues with scale,” says Westwood.
“Like all-flash solutions, converged and hyper-converged infrastructure (CI and HCI) is designed to drastically simplify data centre design and allow the business to be a lot more agile. The other big aspects are integration with next-generation services like containers and the ability to expand into the cloud. They’re fast to deploy and are created to make data centre operation easier and faster to deliver. Good solutions for data centre optimisation, deploying edge services, or trying to simplify the way the data centre is being managed.”
Number three is data centre optimisation.
A big trend is for data centre companies to optimise their existing systems by undertaking airflow management and computational fluid dynamic modelling, which has helped data centre operators create more efficiency and better understand how to design their data centres. Some will work with modular containment around key systems to help optimise airflow. Others might look at new airflow management systems that impact HPC or new converged systems,” says Westwood.
“Similarly, DCIM software and other data centre management tools have been very important and we’re seeing integration with things like machine learning, cloud, virtual systems to leverage management platforms to improve overall data centre functionality.”
And then of course, we spoke about cooling solutions.
“Working with new cooling and power considerations has also been a growing trend. Free or natural air cooling has been a very common way to drive efficiency if the climate is good, rather than running power-hungry mechanical refrigeration or air-conditioning units. As a growing number of data centre service providers build racks and servers that can run in temperatures as high as 27°C, natural air cooling has become a very good option,” says Westwood.
“Gartner estimates that ongoing power costs are increasing at least 10 percent per year due to cost per kilowatt-hour (kWh) increases in underlying demand, especially for high-power density servers. Adoption of liquid-based cooling is increasing, as it is considered more efficient than air-based cooling. The global data centre cooling market using liquid-based cooling techniques is expected to grow at a remarkable rate through 2020. Already, we’re seeing several server and data centre systems being pre-built with liquid cooling solutions. This is definitely something to keep an eye on for the future.”
Westwood believes microgrid architectures will grow in prominence in the coming years.
“Data centre microgrid architectures which are energy systems consisting of distributed energy sources (including demand management, storage, and generation) and loads capable of operating in parallel with, or independently from, the main power grid,” says Westwood.
“These types of architectures supported by data centre management and operation could potentially provide cost savings, emission reduction and reliability enhancement for data centres.”
Westwood says data centre infrastructure management (DCIM) tools and platforms are key for making infrastructure energy-efficient.
“They bring together standalone functions such as data centre design, asset discovery, systems management functions, capacity planning and energy management to provide a holistic view of the data centre, ranging from the rack or cabinet level to the cooling infrastructure and energy utilisation,” says Westwood.
“On-site wind generation or use of renewable energy has been a good option for a number of the larger data centre operators, including Apple, Facebook and Google, who have taken the initiatives to power their data centres using wind energy.
And then finally, Westwood spoke about the Open Compute Project (OCP) technologies.
“Open Compute Project technologies is a change in both Infrastructure and IT architectures which removes centralised capital plant with enhanced hardware into the IT racks, this simplifies architectures, enhances efficiency, is faster to deploy, easier to maintain, and reduces cost,” concludes Westwood.