The OpenDCME Model

The OpenDCME model does not tell you HOW to measure its KPIs. Suggestions are made and experience on what to do and where you could start are on most KPI pages. Most important is that the KPIs are measured and reported to fit your need. Next, identify improvements, plan and implement  them and perform a new measurement using the same method.

Per KPI, a score is achieved. There is no such score as an average-model-score or a total-model-score. And you cannot use the results to compare data centers because every data center has its own unique characteristics. A data center that has primarily network equipment has a different score than a realtime-computing data center. And you can have a very good reason why to score low on a certain KPI. If the ROI to improve that KPI is negative, then accept the low score.

The model consists of four quadrants where each quadrant has its own set of four KPI’s. Here is what each quadrant represents:



Determine the energy consumption of IT and the total data center, the physical space in the data center that is actually used and the air-temperatures at IT and cooling. They answer the question if the facility is in balance and if the cooling system is well tuned.

Click here to read more about the quadrant Facility.


IT Assets

Even with a great PUE/DCiE/EUE/etc there is, with all respect, no reason to believe that IT is doing a good job. What hardware is installed, what capacity is actually used and what is the amount of old hardware? Too much over-capacity consumes energy and therefore costs money. If it is not for “Greening IT” then do it for savings cost on IT hardware, support costs, license costs, etc.

Click here to read more about the quadrant IT Assets.



The previous two quadrant require the input of data. What is the total capacity, what capacity is in use, what capacity demand is on the horizon, how was the capacity used over the last twelve months? Measure to be able to make decisions. Otherwise you are blind and there is no proof that you are in control.

Click here to read more about the quadrant Management.



The result of all previous quadrants together is caused by processes and human behavior. Why did you buy that IT hardware, do you know the energy consumption of those five year old servers, do you consider airflow design when procuring new hardware, ever thought of segmenting the data center into multiple service levels, are the IT and facility team(s) aligned to run the data center at its highest efficiency?

Click here to read more about the quadrant Processes.