understanding energy costs and usage in a data center by Angela Luke
A data center can consume up to 100 or 200 times more electricity than a standard office space, which quickly translates into excessive expenses for continuing that level of energy consumption. As energy costs keep growing, more and more companies are looking for options to keep their costs down, and there are some solutions that can offset these costs while creating a greener IT environment.
IT equipment – the servers and storage – can account for over half of the energy usage at a data center. In other words, most of the energy that is consumed in these centers is a result of just keeping the basics running. A lot of this is because even though the servers spend most of their time running at 20% utilization, they still draw full power most of the time.
According to the U.S. Department of Energy, rack servers remain one of the major power users in data centers, but many of them simply don’t have the option to change, yet, because their offices don’t have the ability to support a different type of server.
Blade servers, for example, can provide much better energy and cooling efficiency, but they require higher voltage circuits for power deliver. A lot of facilities simply aren’t wired to deliver the necessary voltage so the data center has to make do with other solutions.
As the continued move to cloud computing pushes these data centers to keep scaling their offerings, greener solutions that also provide better business efficiencies will help these companies reduce costs and reclaim money, time, and human resources. Taking the steps to use blade servers can give a data center the necessary computing power while reducing the number of redundant facilities and make better use of the available physical space.
There are a range of enterprise solutions that are designed to meet these needs by providing the flexible infrastructure that can deal with many dynamic business situations. Some blade servers, for example, offer power management features that allow users to set thresholds for each blade in the enclosure. They can also provide real time reporting to help understand what is happening in individual blades and as a whole. A simple change like this can potentially minimize the problem of full-powered servers running at less than a quarter of their full availability.
Deploying an effective blade server chassis can potential save a lot of money and energy over the life of the product, but sometimes it is hard to compare the lifetime value of something like new, more efficient cloud computing components to the immediate cost of deployment. However, as energy costs continue to go up, simple changes like these can potentially make a big difference.