Not that I want to spend time doing so, but I would bet that if one performed a search to find out what are the two most over used words related to data centers and technology over the past several years, those two would be “Cloud” and “Green.” Not that these are bad things, nor are they unrelated, but they’ve just been uttered ad nauseam.

Thinking more about the “Green” area, most of what I’ve heard, read and studied focuses entirely on the operating environment of the data center and the back-of-house that conditions it, which certainly is vital, but it also places little emphasis on the IT equipment that actually creates the problem. According to a recent study by Gartner, 12% of the annual data center operating budget is related to energy consumption and that cost is rising by 20% per year. So what if we were able to cut that 12% amount by 75%? In terms of operational issues as well as business drivers, going green depends a lot on the stage in the life of the particular data center in questions and what is the size of the budget that can be applied to the project.

However, benchmarking IT equipment inventory as well as all equipment and designs of the back-of-house is a logical place to start. In order to understand and benchmark efficiency levels of current operating equipment, as I’m sure you are aware, you’ll need to collect or have collected metrics over a significant period of time. Understanding temperature ranges in your data center, hot spots that might exist, the amount of cooling that is being applied and how effectively that cooling is working in the facility are critical pieces of information.

The benchmarking study will probably point to unnecessary redundancies, or under-utilized assets that are no longer providing the availability quotient that was originally intended and can be eliminated or replaced with more efficient equipment. This must be complimented with information about the geography in which the facility is located to further understand and apply any environmental benefits (i.e. free cooling).

However, before moving forward with any information derived from your research and evaluation, I would ask if your IT environment is ready for any server refresh. Along with making any decisions about a refresh, the next item to investigate is the level to which your company has virtualized its server environment. Most companies haven’t taken full advantage of this technology, but this is surely one major step on the road to becoming more efficient and green. Some applications just won’t run well in a virtualized environment, but most will operate to specifications.

If you’ve already virtualized as much as you can and are at the point of being able to refresh a portion of your servers, then this could further promote your green initiative. There are several new server designs manufactured by SeaMicro ( and Calxeda ( that employ chips manufactured by Intel that consume less power, produce less heat and yield almost the same work load as standard blade servers. Typically, the power consumption is about 25% of the typical blade and the heat produced is also approximately 25% that of the conventional blade.

Even if an entire landscape couldn’t be populated with these servers, but they could be used in 50% of the instances that would create a huge power savings as well as a severe reduction in the need for cooling. As part of your IT equipment refresh that will include servers and more virtualization, you’ll likely want to employ more smart power monitoring and other more energy efficient network equipment. Again, the key to continued support of any green initiatives will rely on Data Center Infrastructure Management software, virtualization, installation of newer, more efficient servers (with more efficient power supplies), more efficient UPS systems, DC power (Did I write that? There’s another Pandora’s Box), more efficient switch gear, and standardized processes that encourage sustainability.