Chat with us, powered by LiveChat

Using PUE To Optimize Your Server Room

James Mulvey

Power use effectiveness (PUE) is a metric created by to measure data center efficiency.

Power use effectiveness (PUE) is a measurement method created by as a way of monitoring cooling and energy efficiencies within a data center. Even though PUE sounds like yet another industry catchphrase, it can actually be used as a management tool within the data center.Data Center PUE

PUE Server Room Basics

Gaining insight into the energy efficiency of your server room isn’t always easy, yet by simply dividing the total amount of power supplied to IT equipment by the amount it actually consumes will identify how much power is lost.

Identifying were power is lost is the first step towards making your server room run more efficiently. There are 3 ways in which PUE measurements can be taken:

Level 1

– Measurements are taken from the UPS (uninterruptible power supply), an electrical storage device designed to provide backup power if the main power should fail. It is more like a battery pack designed to kick in as soon as power loss is detected.

PUE LevelFacility power measurements are taken from the data center’s UPS power source at least twice a month.

Level 2

– IT equipment power is measured at the PDU (power distribution unit), which is basically an industrial grade power strip. PUE measurements are taken daily.

Level 3

– This is perhaps the most advanced method of PUE measurement taken continuously throughout the day and may include power measurements from both the PDU and UPS.

PUE Measurements Help Lower TCO

The Uptime Institute estimates that most data centers average a PUE of 2.5 watts, meaning, for every 2.5 watts input one watt is output to IT equipment load. Effective PUE is indicative of a balanced system, everything from cooling to power supply in to power supply out needs to be properly balanced.

PUE measurements will help IT managers identify problem areas within the server room and correct them accordingly. Reducing power inefficiencies in the data center will cut down on operating costs later down the road.

For more information contact

Leave a Reply