The Green Grid Association together with The Data Center Metrics Coordination Taskforce has published a new report, entitled “Recommendations for Measuring and Reporting Overall Data Center Efficiency: Version 2 – Measuring PUE at Data Centers”, which completes the guidelines for applying the Power Usage Effectiveness (PUE) metric. The new information includes specific recommendations for how to measure and calculate PUE in mixed-use data centre facilities.
Back in January 2010 a group of organisations in the US got together to agree on data centre energy efficiency measurements, metrics and reporting conventions. Among others they included The Green Grid, US EPA’s Energy Star programme, US Green Building Council and the Uptime Institute. In July the first report was published, providing recommendations on measuring and publishing values for PUE at dedicated data centre facilities.
The reports came about because while PUE, which determines the amount of energy used by a data centre facility and the IT equipment inside of it, is widely used as a measure of data centre efficiency, the metrics are not always applied clearly and consistently. This is an issue when data centres are compared with each other or over time. The documents set out what should be measured, how and when.
In terms of assessing data centre energy use the organisations agreed the following principles:
• PUE, using source energy consumption, is the preferred energy efficiency metric for data centres.
• When calculating PUE, IT energy consumption should, at a minimum, be measured at the output of the uninterruptible power supply (UPS), although the industry will continue to aim to directly measure the IT load (e.g. servers, storage, network, etc.).
• For a dedicated data centre the total energy in the PUE equation will include all energy sources to the data centre. For mixed-use data centres, the total energy will be all energy required to operate the data centre, similar to a dedicated data centre, and should include cooling, lighting, and support infrastructure for the data centre operations.
The misuse of PUE, particularly in comparing data centres, has threatened to bring the metric into disrepute, so the clarification contained in these two reports is to be welcomed. I doubt it will stop bad comparisons being made, but at least it makes it very clear that it’s not the result of the measure itself.