The rising cost of energy has again put data centre power consumption in the spotlight. Gartner Research has estimated that energy costs, which have traditionally hovered around the 10% mark when it comes to IT budgets, could soon rise to 50% of departmental expenses. This is the consequence of a combination of factors including the rise in energy cost, increased demand for computing capability and continued adoption of high density computing strategies.
Data centres are highly energy-intensive since the servers and ancillary systems such as cooling within them require large amounts of power. With both the number of data centres and their occupancy levels set to rise significantly in future as corporations require greater amounts of IT infrastructure, some providers are starting to take account of their environmental responsibilities.
The Digital Reality Trust, for example, has announced that is to use the Power Usage Effectiveness (PUE) metric as the methodology for measuring and reporting energy efficiency in its portfolio of facilities in North America and Europe. PUE is an emerging standard, promoted by The Green Grid and others in the data centre industry, to provide a simple and consistent method of measuring ratio of power delivered to IT equipment and to the total amount of power used by a facility.
“PUE is a compelling way to measure energy efficiency in data centre facilities because it offers industry an apples-to-apples comparison similar to the miles-per-gallon (mpg) fuel efficiency rating that the auto industry uses,” said Jim Smith, vice president of engineering at Digital Reality Trust. “It will not only provide data centre operators with important benchmarks for measuring and improving data centre efficiency, it will also provide customers with valuable information about how their data centre partners are supporting their green IT strategy.”
For Smith, it’s a case of applying intelligence to the operation of data centres. “Significant gains in energy efficiency can be achieved by designing, building and operating data centres in ways that reduce the overall power needs of the data centre,” he added. “This translates into minimising the amount of power required by non-IT equipment and maximising the ratio of power used by computing systems like servers.”