PUE (Power Usage Efficiency) is a ratio to measure how much more power is required to run IT equipment. It is defined as (Total power supplied to a data center) / (all the power consumed by IT equipment).  This is always more than 1. This number could be made as close as to one but never gets to one or less than 1. This is obvious since we need two types of power to run IT equipment. One is to run IT equipment itself and the other is to run infrastructure to support IT equipment, such as cooling and lighting.

In this blog, the blogger reports on Matt Stansberry’s blog. Stansberry reports on the online seminar of Ken Brill, Uptime Institute. Brill warned that some companies are claiming PUE lower than one. Brill’s point was it is necessary to understand what constitutes the total power. Depending upon how you define the total power, PUE may change.  However, it never gets to be one or less than one.

"Brill said he’s seen companies talking about a PUE of 0.8 — which is physically impossible. “There is a lot of competitive manipulation and gaming going on,” Brill said. “Our network members are tired of being called in by management to explain why someone has a better PUE than they do.”

"If you’re going to compare your PUE against another company, you need to know what the measurement means. “You need to know what they’re saying and what they’re not saying,” Brill said. “Are you going to include the lights and humidification system? If you’re using free cooling six months of the year, do you report your best PUE?”

The blogger also points out a couple of interesting points regarding PUE. The way The Green Grid defines does not indicate PUE being dynamic.  It is obvious that the power consumption by IT equipment and supporting infrastructure changes dynamically over time.  For example, more accesses to servers during the day and less so during the night. Thus, PUE changes even within one day.

Another interesting point is that typical IT equipment, such as server, operates more efficiently when it is loaded up high, compared to a low load. This is because a server consumes power even when it is idle and thus, at a lower load efficiency is not very good. The efficiency ratio improves as more loads are put on the server.  But at a higher load on a server, more total power consumption is observed. In order to conserve power, if making a server more lightly loaded, then it would  make PUE worse.  This indicates PUE alone may not be a right answer to be used as a metric for a data center.

The research on metrics has just started and more work is necessary.