On the same day, two opposing articles have been published with completely opposite points of view.  On one side, Brian Fry argues that location no longer matters and we should build data centers in the most efficient locations and supply fiber connectivity to them.  On the other side, Paris Burstyn argues that location and latency are business critical for companies resulting in Equinix's acquisition of Switch and Data.

At the center of Brian's argument is that low cost, low carbon power is good for data centers.  If you can live with your data centers a few milliseconds away from everyone else's, then your IT operations can be greener and cheaper, a win-win scenario.  But in the world where every millisecond counts, that's not the right solution for all industries.

To say that location is completely irrelevant would be foolish.  When sub-millisecond latency matters in financial trading transactions, locating your data centers across the globe would certinaly put you at a disadavantage.  The New York Stock Exchange (NYSE) has even entered the arena, building their own data center which includes colocation space so that financial companies can be right next to the exchange, all taking place in a building in New Jersey about 10 miles away from the actual trading floor in New York City.  The center will be 400,000 square feet, and operational in 2010.  Not to be left out, NYI is building their own data center down the street to peer with NYSE, and Equinix and Savvis have expanded their data centers.  Even the SEC has started an investigation into the fairness of frequency trading due to the expansion of colocated automated data centers for financial trading.  Sounds a lot like the Office Space /Superman scam, doesn't it?

Financial instituations surely have a latency requirement, but what about the rest of data center use?  Certainly the major Internet service companies, like Google, Yahoo, and Amazon, have put their data centers where power is cheapest (see my blog about The Hype of Green IT).  But they don't serve all their content from one location; they are diversified across the globe to provide a good quality of service to different geographical markets.  So clearly geography does matter, but as long as you're within a "reasonable" distance of your customer (what appears to be on the order of a few hops), the location is less important.  Take a look at the Google Data Center Map, which pegs locations near every major metro area.  Well, unless you're in Australia, the Middle East, or Africa, in which case, sorry but Google is going to be slower for you.  And frankly, why would you locate a data center in the Middle East or Africa, the two warmest climates on the planet. (Although there are on-again/off-again rumors of an Australian data center in development)

Of course, I can't claim the two articles are unbiased.  Decide for yourself, by taking a look at their bios.  Brian Fry is a co-founder of RackForce, a colocation and managed services company.  It makes sense that Brian would argue that his uber-efficient data centers with cheap power can provide a better financial opportunity than one close to other data centers.  Paris Burstyn is a long-time analyst and former IT services consultant, who certainly has seen the ups and downs of data center operations.