If you want to know what I think of predictions read L. Gordon Crovitz's December 29th column in the Wall St. Journal. Crovitz quotes Roman engineer Julius Sextus Frontinus in 10 A.D, "Inventions have long since reached their limit, and I see no hope for further developments." Frontinus's prediction heads a list that Crovitz calls his Top 10 Worst Technology Predictions. Others on the list include Bill Gates, H.M. Warner, Thomas Watson, and Sir Alan Sugar, who predicted that the iPod would be "dead, finished, gone, kaput" by Christmas 2006.
Technology editors are not immune. I'm frequently sent links skewering one editor in the PC space for making light of the computer mouse when it was introduced in 1984. This is why I held off making my predictions for 2010 until the day when one of them came true.
1. We will see increased government involvement in the data center space. On January 6, 2010, Department of Energy (DOE) Secretary Steven Chu announced that the DOE would spend $47 million taken from the American Recovery and Reinvestment Act (ARRA) on data center projects. These funds would support research projects in the private sector and be matched by more than $70 million from large corporations active in the data center space.
In addition we can expect to see government-run data centers start to catch up with their commercial counterparts. Closing this gap will be necessary, if the government is to pursue efforts to further regulate the industry, which it will.
2. End users will continue to pursue a variety of strategies to ease the pressure imposed on their data centers by rising electric costs, increased complexity, and the prospect of cap and trade and other government regulations. They will increasingly turn to colocation and hosting solutions, in spite of the difficulties of raising capital, and cloud computing and SaaS. These changes will drastically reshape the data center landscape
3. Water considerations will become as important as energy in planning and running data centers. It's no surprise that I would make this prediction. Water began to emerge as a recurring theme in my blog as long ago as April. Others realized the importance of this issue long before then.
4. IT and facilities personnel will not learn to communicate, but failures on both sides will lead to all sorts of enterprise problems. Poor designs, construction and/or maintenance will contribute to some major outages on the facility side. DNS attacks, security breaches and network issues will grow in severity. The Department of Homeland Security, which is responsible for protecting civilian computer systems, suffered 850 cyber attacks in the 2 years from 2005 to 2007, and this trend will not abate.
5. Utility performance will cause headaches. Many rate increases will approach double digits, and some will exceed 20 percent. Even so, the nation will experience at least one major outage in 2010. This failure will stimulate Smart Grid development. Increasingly, it will be seen as a no-brainer that the Smart Grid extend all the way to the server level. After making four safe predictions, I'm out on a limb here. I'm also picking the Mets to win the 2010 World Series.
5. Pressure will continue to mount for LEED to issue a certification that is more appropriate for data centers. Climategate does not affect the business case for green and sustainable development. Similarly the world's failure to build consensus at Copenhagen will not slow momentum in this area.
6. Onboard UPS and data center distribution will start to gain traction, and colos, enterprises, and other data center operators will begin to adopt designs developed for truly big and dense data centers. Similarly, more data center operators will make efforts at attaining the kind of super-low PUEs that only operations like Google have attempted. Look for multiple announcements of PUEs below 1.0 achieved by including on-site generation and free-cooling in the calculations.
7. Solutions that made inroads in 2008 and 2009, including containerization and modular data centers, contained rack rows, and other cooling solutions will find numerous new applications, as they will come to be seen as less than exotic. Vendors will improve these products to make them reduce the complexity of these and other infrastructure products. For instance, containerized data center concepts will evolve beyond the crude shipping container retrofits we have seen so far.
8. Not only will Data Center Pulse announce the results of its Chill Off 3, but one or more industry groups will apply this model of performance-based testing to other data center systems.
9. PUE will gradually give way to another, more holistic, data center metric that will address compute performance, capital and efficiency optimization and TCO.
These predictions were developed with input and guidance from other famous prognosticators, including Ebay's Dean Nelson (also the Data Center Pulse), HP Mission Critical Facilities's Peter Gross, Gilbane's Dennis Cronin, Power Management Concepts’s Peter Curtis, and Jacobs's Kevin Dickens.