As consumers, we want more content, more information available and continuous on demand support. In the commercial sector, companies want to capture business intelligence, deliver demanded information and services and manage a mounting volume of data. As a result, the amount of data center power that is being consumed is rapidly growing.
A recent New York Times blog evaluated this growth, highlighting that some of today’s data centers contain hundreds of thousands of computer servers managing tasks as diverse as posting pictures to Facebook (News - Alert) to complex physics problems. The consumption of data center power has grown into a primary concern with activities it aims to support. According to a recent industry census, electricity use is expected to grow by 19 percent in 2012 alone.
In reality, this growth, even if it is primarily a result of increased data center power, is actually less than users have seen in years past. It does indicate there is no less demand for support in the cloud and servers that are accessible by millions of users. Given recent industry projections the demand for corporate and private clouds is only expected to continue to grow. The good news is that the management of big data centers is getting more efficient as engineers continue to re-examine every aspect of production and the consumption of data center power.
According to Stephen Hassell, head of data center infrastructure management for Emerson (News - Alert), a St. Louis-based engineering company, cloud companies such as Facebook, Google (News - Alert), Amazon or other big telecommunications companies will manage their own server designs for such environments, yet they are still seeking ways to identify reductions in the consumption of data center power. Hasseell’s company works with data center managers to deliver the more efficient use of data center power.
One early solution leveraged by a number of companies to reduce data center power use was virtualization. This technology platform enables one server to take on the work of several and raises the individual power efficiency of a single box. With this platform, companies can focus on the data center power use of individual computer applications as tasks can vary in terms of complexity and the amount of power consumed. With another layer of computing inside the system, data center managers can allocate which applications can run more efficiently on specific computers.
Our reliance on data center power is not expected to lessen, demanding that managers in this space identify efficient ways to manage increasingly levels of data and complex applications to support the demands of the industry.