If we ask any IT organization about its budget, the common answer you hear is that the budget is cut by XX% for the next fiscal year. At the same time, IT services expected from the IT organization are always on the rise. In addition to the standard OPEX and CAPEX an IT group was responsible for previously, they are also under pressure to utilize the energy efficiently and reduce carbon emissions. This means that the IT group is now forced to monitor their energy utilization and make sure the energy efficiency is high so the operational cost can be lowered. Due to this fundamental change in addition to the rising demand for higher energy efficiency, the need for green datacenters is increasing.

A green datacenter is one in which the energy efficiency is maximized and the environmental impact is minimized. It’s a well documented fact that datacenters are major consumers of energy. Additionally, energy equipment such as PDUs, UPSs, etc. and the cooling units contribute to energy loss.

So, how do we maximize energy efficiency? We must rely on industry standard energy efficiency metrics – such as the Green Grid’s Power Usage Effectiveness (PUE), which is the ratio of total power coming into the datacenter to the power used for IT load and Data Center infrastructure Efficiency (DCiE), the reciprocal of PUE in percentage value, or the Uptime Institute’s Site Infrastructure Energy Efficiency Ratio (I-EER) and IT-PEW (IT Productivity per Embedded Watt).

One way to maximize energy efficiency is to adopt one of the standard metrics previously mentioned, and based on that, continuously monitor your energy utilization. Once you understand the energy utilization and associated losses completely, you can determine what changes would most likely help save you both energy and money, and then implement those changes. For example, if a server’s operating temperature is from 0ºF to 120ºF, then you don’t need to run the datacenter at 60ºF. I saw a stat recently that stuck in my mind – it noted that for every 1ºC rise in the datacenter temperature will save nearly 4% energy cost – that’s big savings when we’re talking about typical datacenter energy costs for many large enterprises. Implementation of advanced cooling techniques and economizers can help save a lot of energy and cost associated with cooling.

If we consider PUE as our energy efficiency metric, then PUE is directly proportional to the total power coming to the datacenter and inversely proportional to the power drawn by the IT Load. PUE value should be as low as possible with an ideal value of 1, which means that all incoming power to the datacenter is being used for IT Load. If a datacenter has a PUE of 3 (which is true in many cases), that means only 33.33% of the incoming power is being used for IT load. In other words, if an organization is spending $1M for their datacenter energy, then only $333K is used for the primary objective of IT load/equipment. Remaining $667K is either related to cooling or energy losses. Once we know these numbers, then we can implement necessary changes, best practices, and guidelines to improve the energy efficiency.

Now that we’ve gone over the basics of energy efficiency metrics, in a future post, I’ll share some tips for how best to maximize energy efficiency, helping to move toward a greener datacenter.

Image used under Creative Commons License.