The 100 MW Data Center
Or, the data center side show.
Apart from the hyperbolic — not to mention misleading — headline, is there really anything to see here? We all know that these guys are building facilities that are tuned to their very specific and unique requirements, but other than giving a PR flack an opportunity to put out another press release, does anyone really care about these guys’ media infatuation and grappling to prove “whose is bigger than whose?” While I respect what these companies have done to date, I don’t think that the real value of their efforts is to serve as some data-center-on-steroids freak show. Maybe it’s only me, but I fail to see the fascination with these artificial constructs.
I realize that we will always have a fascination with size, but what exactly was the purpose of this article? While a 100 MW facility is certainly huge — actually that’s more like a nuclear power plant with racks. Now that would be interesting — does it really mean anything to the average financial service company that needs a new 2 MW location? Not really. If you were to only read the title of this article, you’d think that one of these behemoths was soon to be coming to a neighborhood near you. However, since data centers the size of small planets aren’t about to invade suburbia any time soon — talk about your zoning issues — this really isn’t a burning industry issue.
Maybe these guys are actually getting a little embarrassed by all this misplaced attention. A number of them have started releasing video tours of their facilities. I guess this is to make us feel that deep down they’re still just regular data center guys. “Welcome to our humble computer processing abode. We’re just like you only we use enough electricity to power Danbury, Connecticut.” I must admit, when you watch these high-tech versions of home movies, they pretty much look like your average industrial plant except for the fact that Segways are needed to get around. The downside of this focus on size-related milestones is that, through their efforts, these same firms have delivered a great deal of practical value to the industry as a whole.
Many of the functions that these “large scale” data center companies had to develop to achieve their ability to support vast amounts of data storage and processing capability established methodologies are now used in many enterprise facilities. For example, the ability of many data centers to deploy load groups ranging from hundreds of kW to MW is a direct result of the need to quickly implement capacity to address large volumes of demand that cannot not be accurately forecasted. This need for rapid large deployments has also manifested itself in modes of performing moves/adds/changes that have been widely embraced by the data center industry, albeit on a smaller scale.
In a sense, these efforts of these companies have served as a de facto R&D lab for the industry as a whole. Operating in an environment largely stripped of “mission critical” fear due to the fact that their software can bear outages, these future members of the 100 MW club have not only developed new modes of operation but have forced changes in the hardware used to operate enterprise level facilities. By continually pushing the envelope on operating temperature and humidity, they were able to work with standards organizations like ASHRAE and put pressure on equipment manufacturers to expand their own parameters to deliver warranties to support these new capabilities.
Maybe this emphasis on size-related milestones is the industry’s version of the shiny bauble. With the cloud stuck in neutral and big data a concept everyone loves but few understand, things like 100 MW data centers provide us with concepts that are easy to grasp and rally around. Don’t we all want to be the biggest and best at something? Unfortunately, this fixation on size leads of to overlook the practical value of the more “mundane” innovations of these same companies. In some ways doesn’t this harken back to the space race of the ’60s? While going to the moon was the prize that we could rally around, the byproducts of this effort (Teflon, anyone?) were what delivered real and practical value to us.
Will the relentless pursuit for the 100 MW data center continue? Of course, it will. We can also assume that we will see more and more providers proudly touting the construction of their latest mastodon-grade facility. However, these achievements will hardly be harbingers of the beginning of a new epoch. The real gains in our industry will come from the efforts and developments that arise to underpin these size-centric achievements. Not as captivating or noteworthy as breaking the 100 MW barrier perhaps, but probably much more important.