Business Continuity Planning for Data Centers and Systems: A Strategic Implementation Guide

Chapter 5. Data Centers Growth Velocity

I have already discussed the data center of the 1980s and 1990s. During that time, I effectively predicted 30 to 70% growth over 10 to 13 years. That means if users needed (day 1 and future) 100,000 square feet of white space, they would plan for 130,000 to 170,000 square feet total (inclusive of environmentals to support the white space). That figure was based on existing velocity of growth for white space. Power densities were 15 to 25 watts a foot, and cooling was fairly static. Moore's Law, although in place at this time, did not anticipate the more powerful chips associated with cooling configurations until years later. More recently, particularly following the phenomena of Y2K and the terrorist events of September 11, there has been accelerated growth and interest in large data centers. The "mega-data center" had lost its place (90% of all data centers are less than 10,000 square feet). The cost and time required to design the 50,000 to 100,000 square feet of white space was and is overwhelming. It often made sense to augment human space or office space within a strategic asset rather than taking the time and expense to secure, design, improve, and maintain a large data center. Operating expenses over 15 to 20 years are staggering, and are the justification many users give for not designing a large data center. Data centers are not their core competency or core business; why should they pour $3 to $10 million a year into just the maintenance ...