DCIE or Data Center Infrastructure Efficiency was developed by the Green Grid as a benchmark for energy efficiency in the data center. DCiE’s reciprocal is PUE which stands for Power Usage Effectiveness.

The formulas for both DCiE and PUE are as follows:
PUE is defined as follows: PUE = Total Facility Power/IT Equipment Power and its reciprocal, the DCiE is defined as: DCiE = 1/PUE = IT Equipment Power/Total Facility Power x 100%

According to the DOE, The applicants who are interested in conducting research and receiving funds must show a plan for the technologies to be demonstrated and the adoption of other best energy management practices to improve a data/ telecommunication center’s energy intensity performance (energy consumed for a given level of useful computational work) by more than 25 percent and have a Data Center Infrastructure Efficiency (DCIE = IT energy / total facility energy usage) of 0.80 or greater.

So you may ask yourself, it sounds like the DOE has a game plan and a benchmark they wish to follow from here on out. Yes, they do have a plan, but many in the industry are not convinced that PUE or DCiE are the best metrics to use to measure efficiency.

The argument has many sides. One point notes that as the PUE # gets higher the worse things are—which is counter-intuitive. A better argument posed by David Cappucio, a member of the Gartner Blog Network states: “While these metrics will provide a high level benchmark for comparison purposes between data centers, what they do not provide is any criteria to show incremental improvements in efficiency over time.  Most importantly, they do not allow for monitoring the effective use of the power supplied – just the differences between power supplied and power consumed. There is no negative, or positive effect of more efficiently utilizing the compute resources at hand, so a data center with a PUE of 2.0, running all x86 servers at 5% utilization, could have the same PUE as another data center running the same number of servers at 50% utilization, effectively producing 10 times the compute capacity as data center 1.”

In an interview with Peter Gross, vice president and general manager for HP Critical Facilities Services, he notes that “HP Critical Facility Services has been using these metrics, but it is fundamentally wrong. It is very difficult to measure all the factors as well as do a fair assessment comparison between facilities. How can you compare a data center in Houston with a data center in Iceland?”

Now that the Department of Energy has adopted the DCiE metric what happens next? The next logical scenario may require US data centers to prove that there data center meets a certain DCiE level or risk losing carbon credits if the carbon system is adopted in the US.

The DCiE is an interesting move by the DOE, but are there other choices? Yes, there are.

In Mr. Cappucio’s blog he notes the following:  “A more effective way to look at energy consumption is to analyze the effective use of power by existing IT equipment, relative to the performance of that equipment.  While this may sound intuitively obvious (who wouldn’t want more efficient IT).  A typical x86 server will consume between 60% and 70% of its total power load when running at low utilization levels.  Raising utilization levels has only a nominal impact on power consumed and yet a significant impact on effective performance per kilowatt. Pushing IT resources towards higher effective performance per kilowatt can have a twofold effect of improving energy consumption (putting energy to work), and also extending the life of existing assets through increased throughput.

If major IT assets were evaluated in this manner it becomes clear that not only can more efficient environments be created, but that individual asset utilization levels can be increased, effectively improving the performance per square foot within the data center, and potentially deferring the construction of a new data center.”

There are other metrics that are also available, but it appears for now in the eyes of the DOE and the research grants they will be offering that DCiE is the benchmark de jour.

Comments (0)

Subscribe to this comment's feed