IT System Efficiency in the Data Center

In the past, IT equipment manufacturers and their data center operators were focused more on maximum performance, not energy efficiency. While the newest generations of computing hardware have continued to increase in performance, they have also become highly focused on energy efficiency. This article is the second installment in a five-part series on energy efficiency.

Julius Neudorfer

August 21, 2012

4 Min Read
DataCenterKnowledge logo in a gray background | DataCenterKnowledge

In the past, IT equipment manufacturers and their data center operators were focused more on maximum performance, not energy efficiency. While the newest generations of computing hardware have continued to increase in performance, they have also become highly focused on energy efficiency.

In fact it has been shown that in many cases the cost of energy for an older commodity server is higher over a three year period than the cost of the server itself. This is especially true in older data centers when the cost energy for the supporting power and cooling infrastruc¬ture is added to the total energy cost (See Understanding PUE in the facilities section of the DCK Guide to Energy Efficiency, page 5). This article is an abridged version of the IT System Efficiency section of the DCK Executive Guide to Energy Efficiency and the second article in a 5 part series on Energy Efficiency.

To help simplify and comparatively quantify the hardware aspect of energy usage and efficiency and promote this concept, the U.S. Environmental Protection Agency (EPA) instituted an “Energy Star” program for data center IT equipment. The Energy Star program initially introduced the first version of the Server specification in 2009 and continues to update and expand its list of included equipment. Currently, they are also working on Energy Star for Storage and Networking equipment specifications and it is expected to be released later this year or early 2013.

According to the U.S. EPA “Computer servers that earn the Energy Star will, on average, be 30 percent more energy efficient than standard servers.” However, it should be noted that in some cases the energy requirements for Energy Star rated servers such a the widely used “1U” volume server, (with a single CPU, one hard drive and one power supply) can be substantially better than the average figure of 30%. In many cases it could use as much as 80% less energy, when compared to a comparably equipped 2–3 year old typical “commodity” server. As such, an IT hardware refresh can significantly reduce the total energy required in the data center, and in fact may offer a very short ROI time frame.

Hardware alone is only one part of the IT energy efficiency factor, the software (Operating Systems and Virtualization, as well as the myriad of standardized and custom Applications), are also critical factors in determining and improving the overall computing performance and therefore efficiency, as well as total energy use. In the past, there were high numbers of individual applications running on individual servers, which were not constantly or highly utilized. However, these servers typically used 60–70% of the full load power while sitting idle.

The introduction of virtualization software and server consolidation has helped to improve the ratio of idle servers drawing significant power with little not productivity. Moreover, one of the important features of Energy Star servers is active power management, which significantly reduces the power the server draws while idle.

Furthermore, the advent of the so called “blade server” which is a central chassis with redundant shared power supplies (which improves energy efficiency), that can hold many individual server “blades,” which allows packing more computing power into a smaller space. The blade server has also helped drive server consolidation and virtualization projects for many organizations.

The entire IT industry is now driven to improve energy efficiency. The IT hardware manufacturers have recognized that not only do they need to inherently improve the energy efficiency in their products, they are well aware that the largest use of energy in the data center facility itself is used by the cooling systems. They have worked to make their hardware more robust to allow it to operate at higher temperatures, which in turn reduces the cooling systems energy requirements.

So what does this mean to the data center facility and it cooling system design and operation? Data centers have historically kept very tight environment conditions to help ensure the reliability of the IT equipment. This was originally driven by older equipments susceptibility to temperature and humidity changes as well as a very narrow range of “recommended” environmental conditions mandated by the equipment manufacturers themselves. While there IT equipment has clearly become more robust, the data center industry as a whole is more conservative and has not fully taken advantage of the cooling energy saving opportunities the new equipment offers.

A final note on IT energy usage; based on all these improvements, one would tend to assume that since each generation of IT equipment offers greater efficiency that the overall energy usage for IT systems would begin to decline. However, an insatiable demand for more features, applications and more data that is be generated, read and viewed by both consumers and businesses keeps IT system expanding, ultimately, driving up the need for more power for IT equipment.

For the complete article on IT System Efficiency download the Data Center Knowledge Executive Data Center Energy Efficiency in a PDF format compliments of Digital Realty.

Julius Neudorfer is the CTO and founder of North American Access Technologies, Inc. and writes for Data Center Knowledge on issues and strategies relevant to senior business executives.

Subscribe to the Data Center Knowledge Newsletter
Get analysis and expert insight on the latest in data center business and technology delivered to your inbox daily.

You May Also Like