So far, many believe that a data center is only a place to store digital data. In fact, a data center also consists of all components necessary to power and run the entire facility. In a cloud computing system, the virtualized infrastructure consists not only of the storage space, but also the memory, the processor, and other computing components that are necessary for running the entire computing and network system. The performance and efficiency of a data center is thus highly affected by the performance of those physical components and how well the facility accommodates them. Because most companies are now relying on data centers to manage their network system, properly measuring and calculating the performance and efficiency of data center is essential to make sure that the data center and all components within will be reliable for a long run.
Two Ways to Measure Data Center’s Performance and Efficiency
When concerns relating to data centers’ usage of resources become apparent, The Green Grid Consortium announces that a metric called PUE (power usage effectiveness) will be used to measure data center efficiency and to determine whether a data center operates with perfect performance and efficiency. PUE is a ratio estimating the total power that a data center receives divided the power used by the data center’s IT equipment. Ideally, PUE should be 1.0, which determines that no overhead power is within the system. However, PUE 1.0 is entirely hypothetical and a zero overhead power situation is a very unlikely situation. As long as PUE ratio is below 2.0 and close to 1.0, the data center operation is considered efficient. Today’s PUE standard determines the following assessments.
Another metric introduced by the Green Grid consortium is called DCIE (Data Center Infrastructure Efficiency). DCIE is a percentage calculated by dividing IT equipment power with total facility power. Because it is a percentage, the calculation result is multiplied with 100%. The nearer the measurement result to 100%, the more efficient the data center is considered.