I've just returned from a roundtable discussion with Oracle's president Charles Phillips in which he mused on the current state of the software giant in a manner typical of a man who has spent more time than is healthy undertaking media training – ie he delivered a series of supremely articulate responses while saying very little of any actual interest.
However, one response did jump out which served to highlight the extent to which environmental issues continue to remain a niche concern for vast swathes of the IT industry.
Phillips was asked the now de rigeur press conference question about what the company was doing to limit the environmental footprint of its products, to which he rather dismissively responded: "There are only so many things we can do as we don't make the things that use the power."
As one hack pointed out this is more than a little disingenuous because a huge number of the machines that do use the power are only plugged in so that they can run Oracle's assorted databases and applications.
Perhaps a little chastened, Phillips said that its new 11g database had two "green features" in the form of advanced partitioning and compression, both of which limit the energy and costs associated with data storage.
However, despite Oracle's latest innovations it was Phillips' initial response that appeared to more neatly capture where energy efficiency sits in the list of Oracle's priorities.
He is obviously right that there is only so much Oracle and other software vendors can do to limit IT's gargantuan energy footprint and that the bulk of the responsibility for tackling the energy crisis must sit with the hardware vendors.
But software developers still have a huge role to play in creating solutions that require less computing power and less storage, even if it is an approach antithetical to their entrenched desire to develop software that strains at the limits of the hardware's capabilities.
I'd be intrigued to know how many developers at Oracle - and SAP and Microsoft come to that - have been instructed to consider energy requirements when they are designing software. My guess is not too many.
Unfortunately though there is no way hardware vendors can maximise improvements in energy efficiency unless software developers begin to consider the power implications of their work. Few of them would say it out loud for fear of riling their biggest partners and jeopardising one of the main reasons customers instigate hardware upgrades, but they will need far greater co-operation from the software sector than they are getting at the moment if their green aspirations are to prove successful.
It is not a perfect analogy, but the stance summarised by Phillips is not unlike a designer working on a car chassis insisting that there is not much he can do to improve energy efficiency on the grounds that it is improvements to the engine that are really required.
With IT, as with all other industries, truly green business models are only possible through holistic development that incorporates everyone involved in the process. The sooner Phillips and his counterparts at the world's biggest software companies realise this the better.
