Is efficiency all it’s cracked up to be?

By Zahl Limbuwala, CEO at Romonet.

  • Monday, 18th August 2014 Posted 10 years ago in by Phil Alsop

According to the analyst firm Ovum, businesses are now willing to spend more on trying to enhance the energy efficiency of their data centres. Yet, while aiming for better efficiency within the data centre is all very well, organisations need to be careful that they aren’t focusing on efficiency at the detriment of TCO. Essentially, organisations need to understand the total cost of ownership of their data centre when considering efficiency improvements because all data centres .To do this, businesses need to be certain they are addressing the most basic concerns: ensuring that their data centre is operating correctly and as the designer intended it to operate. Whether planning a new data centre from scratch, or investing in an existing site, understanding the influences on performance and cost is crucial.

Understanding the costs of your data centre
At the most basic level, many businesses are still spending significant amounts trying to increase efficiency and reduce costs within the data centre, without fully understanding that every data centre has a tipping point at which improved efficiencies can no longer be made within the company’s expected Return On Investment targets. Key to resolving this is being able to not only measure but also predict the performance of the data centre: as this helps identify exactly how the data centre costs will respond to changes. Without this ability to accurately predict and then measure the financial impact of every potential decision or change, organisations are not able to identify where that tipping point is for any given facility. This, in turn, prevents data centres being run more like a business asset, where costs are fully understood and any changes to the use of that asset have an understood cost impact. A full understanding of their costs can prevent data centres becoming a black hole on IT departments’ balance sheets, sucking in investment with little evidence of return.


Understanding these costs goes far beyond simply considering the IT hardware installed in the data centre. In today’s modern data centre 80% of the TCO is locked at the time it is constructed. Minimising mistakes’ at the design phase will have a profound effect on the data centre’s cost-effectiveness over its lifetime. As a result, businesses must understand the precise financial impact of the choices open to them at the time of design and what if any impact they can have during its operational life. There are already best practices for designing and building a data centre that should help organisations ensure they are put together in the most cost-effective manner possible. For instance, the European Commission’s Code of Conduct for Data Centres aims to assist by applying a series of best practices to the physical design of the facility, as well as the devices and software that operate within it. Again, accurately predicting the impact of any decisions at this stage will help create the most cost-effective data centre possible.


Usage and cost
Whether planning a new data centre or investing in an existing one, businesses need to be certain that their expectations, and predictions, are realistic. For instance, since the majority of energy costs come from the overall building and design of a data centre, one operating at full capacity will be far more efficient than one only running at 60% or worse. In the real world, the vast majority of data centres will never reach 100% capacity; yet efficiency is still calculated and predicted using full-capacity figures. By predicting costs based on a more realistic figure, such as 50% capacity, businesses can be certain that they know the level of performance they should actually receive, rather than a theoretical maximum.


This level of understanding needs to extend throughout a data centre’s lifetime. Reinvestment will be necessary over a data centre’s existence; yet knowing whether an investment will help increase performance and reduce costs in the long run is often more a matter of assumption than prediction. Cooling is a prime example of this: on paper, upgrading a data centre’s cooling systems to more efficient technology can seem like a straightforward way to increase data centre efficiency. However, once the costs of replacing technology and actually powering cooling systems themselves are considered, the actual savings may be far smaller than expected; or even non-existent.


Key to all of these decisions is being able to predict the impact and total cost of any potential data centre investment. Armed with this knowledge, businesses can make decisions that reduce the energy use and cost of their data centres even if these decisions appear, perversely, to make the data centre less efficient. Ultimately, businesses need to prioritise the total cost of ownership of their data centre over concepts such as efficiency. Efficiency is only ever a means to an end; the most efficient data centre in the world is worthless if the TCO is still higher than the business can afford or that of your competition. By understanding and predicting the TCO of their data centre, businesses will be able to reduce costs and increase efficiency as a matter of course.