Checkmark=facebooklinklinkedinOngoing indicator +Left-facing arrowRight-facing arrowCheckmarkDown-pointing chevronLeft-pointing chevronRight-pointing chevrontwitter

Comparing our energy efficiency to datacenters using PUE and SPUE

At Nerdalize, we want our energy consumption to be as low as possible, which means it is essential to use our energy efficiently. In order to do so, we find it important to be able to calculate our energy efficiency as accurate as possible. In this blog, you can read about how we are able to calculate our energy efficiency, using the PUE and SPUE score, and how this allows us to compare our own efficiency with other datacenters.


How to calculate the energy efficiency of the cloud

A common method to determine the energy efficiency of the cloud is by calculating the power usage efficiency (PUE) of datacenters and the server power usage effectiveness (SPUE) of their servers. Although we do not need to build an entire datacenter, the PUE and SPUE scores can be used to calculate the energy efficiency of the Nerdalize servers and our network infrastructure. This makes it possible to compare our distributed cloud services to the cloud services of a datacenter.

The lower the PUE, the higher the efficiency of the facility.

The lower the SPUE, the higher the efficiency of the servers.

The PUE score is the energy efficiency of the entire facility, for example the entire datacenter or the network of Nerdalize CloudBoxes. However, only using the PUE to compare the datacenter efficiency, does not take into account the energy efficiency of the servers themselves. We incorporate the SPUE score in this comparison, through which the efficiency of the servers can be calculated. For the calculation of the SPUE, the useful IT power is used. Useful IT power is the power that is actually used by the hardware to perform computations. This does not include the power loss at server level and the power used by supporting equipment like fans and switches.

PUE = Power usage of the data center divided by Power usage of the servers, SPUE = Power usage of the servers divided by Useful IT power

The Nerdalize CloudBox vs. datacenters

In our blog on life cycle analysis, the total amount of CO2 emissions per KWh of useful IT power is calculated, where both embodied and the operational emissions are taken into account. The operational emissions are caused by the energy use. The energy efficiency is determined by calculating the PUE and SPUE scores. For the different scenarios the energy efficiency is expressed by their PUE and SPUE:

A traditional datacenter

An average datacenter in the Netherlands has a PUE of 1.7, a SPUE of 1.6.

A high efficiency datacenter

A high efficiency datacenter has a PUE of 1.17, a SPUE of 1.25.

A Nerdalize server-header

The Nerdalize CloudBox has a PUE of 1.02 and a SPUE of 1.07.

Since the Nerdalize CloudBox only needs a small amount of energy, which is only primarily by the servers, the operational emissions are very low.

How our PUE and SPUE can be so low

What makes our energy use so efficient, is that almost all the energy we use goes directly to the computations performed by our servers. Because of our efficient cooling system, only 2.2% of the total power consumption of our servers goes to cooling and general overhead. Resulting in a PUE as low as 1.02! Of the resulting power, 6.7% is lost on the power supply unit and the network switch, which makes that we have a SPUE of 1.07.

This efficient system makes it possible to use only the minimum amount of energy. Resulting in a cloud solution with a very low carbon footprint. Because we save on energy and don’t have the costs of building a datacenter, our computing power is also very affordable. Up to 50% more affordable than other cloud providers!

About our CO2 savingsheater
Liesanne Wieleman
posted this June 1