Some of the biggest costs for data centers come from overhead energy consumption, and a great deal of that energy goes toward cooling the centers. In fact, some statistics show that cooling systems take up about half of a data center’s energy intake. Really large data center operators can afford to run off of super-efficient designs, but what can smaller centers do? Inefficient cooling is a big problem for many smaller data centers.
Small Data Centers Exist in Every Community
Visit a university campus IT center or a local government IT facility. These smaller centers don’t get a lot of attention, but they are home to a large amount of the world’s IT equipment. This also means that the smaller facilities are responsible for a big chunk of the energy used up by the data center industry. It follows logically that inefficient cooling remains a problem for the entire IT industry.
The Main Problem
The problem then is that these smaller data centers are operating with inefficient cooling systems that impact the entire industry, but they don’t have the resources to significantly improve. The small-town university IT department doesn’t have room in the budget for major infrastructure upgrades. Further compounding the problem is the fact that some of these data center teams don’t even see the energy bills and remain unaware of their role in this problem. Without that awareness, there’s no motivation to make an effort to reduce energy consumption or improve cooling efficiency.
Hot Spots and Redundancy
The heart of the problem is that too many data centers are being overcooled. This happens for two reasons. The first is that hot spots must be treated, and the rest of the center is overcooled as a byproduct of that intense cooling – basically a result of improper air management systems. The second reason is redundancy. This preventative step is necessary in any data center, but with the same solution (improved air management), inefficient cooling can be reduced.
The Search for a Solution
A solution has already been determined: Simply install the proper controls and increase knowledge of the centers’ actual cooling needs. This will keep redundant units in standby mode, only kicking them on when they become necessary. Sadly, too many smaller data centers don’t have those systems or the resources to implement them. Reliability must be the top priority for data centers, but the development of efficient cooling practices must become more important, because without some improvement from these smaller data centers, the entire industry will continue to be plagued by this problem.
