A slow, steady decline in the number of data centers in the U.S. that was spurred by the economic crisis in 2008 will continue over the foreseeable future, the research firm stated. There are approximately 2.94 million data centers nationwide, a number which is anticipated to drop to 2.89 million by 2016. In terms of sizes, however, total data center space is expected to grow to more than 700 million square feet in 2016, an increase of nearly 90 million square feet from present levels, according to the forecast. In four years, IDC also expects service providers to grow to take up more than one-quarter of all large data center capacity across the U.S.
The biggest impacts to data center changes in the U.S. has been server virtualization and consolidation, which IDC VP of Data Center and Cloud Research Richard Villars said has led to the “dramatic” reduction in the need for smaller, internal enterprise servers. Even as virtualization has cut back on smaller server efforts, the rapid expansion of data volumes and business prospects has led to the reliance on these larger storage initiatives. Scale, performance and centralization efforts on growing data volumes rely predominately on mid-sized and large data centers, explaining IDC’s prediction of greater data center sizes. And, in the U.S., efforts with those larger data centers have been increasingly farmed out to offsite storage centers and cloud deployments, especially when enterprise data managers factor in other aspects like energy costs.
“There is value in not being in the data center business if you don’t have to,” says Villars.
IDC is in the midst of preparing global looks at the footprint of data centers, though Villars says the U.S. continues to lead in the decline of smaller servers and data center spots, with China following a similar model. Even with cloud storage popping up worldwide, there are restrictions to cross-border deployments across Europe and organizations in Asia outside of China, along with the “desire to still possess the equipment” required for enterprise data storage, Villars says.