I wrote and posted on my blog, “Is IT a utility?“, it’s unfinished and there’s some dubious stats and treatment and so I have created this supporting article. I created a montage of the coming of the water network which deserves some words. They key insight is that power transmission loss is a source of economies of scale if we build p2p energy generator systems, energy is enabling P2P generation, and some of the early distributed computing solutions such as SETI did this for IT in the home.
The inability to store electricity and compute cycles is an interesting common differentiator. I did not explore the fact that when networking bits, order is important whereas the order of the ‘cc’ of water or electrons is not. I did not use a hammer to explore Nick Carr’s idea that there are ‘infrastructure industries’ with which I disagree, and my recent insights ask if the power transmission cost and the centripetal force of data mitigate against the economics of the megadata centres.
Another unfinished thought is that if we consider future customers to be queueing for the unused capacity, then Erlang’s theory says that it’s most efficient to have one queue; does this imply one data centre? I think not but this is why Gateway experimented with ‘rooms’ for technical support.
These are basic economics arguments that reinforce the jurisdictional constraints on system location.
I have bookmarked my web sources, and here’s my “energy” tagged bookmarks
The World Bank’s numbers are after transmission loss.