I subscribe to Bill St. Arnaud's CA*net 4 News mailing list (he's got a blog too). Bill works at CANARIE Inc. which operates CA*net, the world's first national optical Internet research and education network (est. 1998 ) in Canada. At first I subscribed to learn more about different models of fiber ownership, specifically home-owned fiber, although now there seems to be a shift in his interested more towards SOA, cloud computing and zero-carbon IT services. Which is also interesting of course. From a look at his blogger profile he has a host of other interests, like in e-democracy.

His usual thesis is that IT can go zero-carbon by locating virtualization and cloud commodity services at places like this. Impending regulations like cap and trade are thought to provide economic incentives for this (there have been many criticisms of these systems btw). Today Bill sent out an article that appeared on Slashdot as well that refers to a UC Berkeley research paper paper called "Above the Clouds: A Berkeley View of Cloud Computing" (pdf) which corroborates his thesis on economical terms. The paper's main task is listing the main obstacles to cloud computing success, which, according to the researchers, are:

• Availability of service
• Data lock-in
• Data confidentiality and auditability
• Data transfer bottlenecks
• Performance unpredictability
• Scalable storage
• Bugs in large distributed systems (author's note: anyone working with grids knows how true this is)
• Scaling quickly
• Reputation fate sharing
• Software licensing

As Bill says in his posting, reliability is a problem for zero or low carbon emission cloud services, since cloud computing must be geographical distributed to be reliable, but you can't find low-emission-and-cost energy everywhere.

Here are some some figures about energy production cost that Bill cites from the paper: