Smart thinking about green computing could make a big impact, says Bill Thompson.
Mining for minerals used in mobiles is endangering gorillas
Like every other product of the advanced manufacturing capabilities of a long-industrialised society the computers that surround us - and, for the pacemaker wearers among us, that we have taken into our bodies - carry an environmental cost.
Silicon may be cheap, but turning it into processors requires vast amounts of energy, clean water and many potentially toxic chemicals.
Some of the raw materials used elsewhere, like the coltan in our mobile phones, are extracted at great human and environmental cost.
Displays and casings may contain heavy metals and damaging chemicals, while the disposal of old computers is becoming a significant issue.
And the billions of processors, hard drives, screens and network devices that we increasingly rely upon consume more and more electrical energy, much of it wastefully generated from non-renewable sources that release carbon into the atmosphere.
Faced with overwhelming scientific evidence that human activity is behind the current shift in the world's climate and that both temperature and sea level will rise significantly in the next hundred years it is clear that we must take action.
Taking small steps in our daily lives is clearly important, both because the cumulative impact of millions of people turning off their big screen TVs matters and also because awareness of the issue will drive real political and technical progress.
But turning off the computers and dismantling the modern world is neither feasible nor desirable, although if we don't start telling our politicians that we expect them to believe the science and offer new policies then we may not have the choice.
Andy Hopper, professor of computer technology at Cambridge University, is heading up a team of researchers exploring how we can use the technology to solve the problems to which it contributes.
Under the banner "Computing for the Future of the Planet" the Computer Lab has outlined an adventurous programme of research into the ways in which computing technology could help us maintain our current way of life, and even give more people access to the comfort, safety and pleasures that characterise modern life in the West.
And while it may not please those who believe that we must somehow suffer for the sin of making the most of what the planet has to offer, what they have done is to take the same sort of engineering approach to the problem of building a sustainable and supportable economy that the great Victorian engineers took to the issues of their time.
Only this time questions of environmental impact, energy efficiency and social justice are part of the calculation.
Prof Hopper and his team, lead by Andrew Rice, Alastair Beresford and Robert Harle, are asking some interesting questions, and they may come up with some interesting answers.
After all, the Cambridge Computer Lab has something of a reputation for innovation, having been home to EDSAC, one of the world's earliest computers.
They are looking at whether digital alternatives to physical activities really make a difference, and whether the environmental cost of creating the iTunes economy is actually larger than the CD-based one that preceded it.
They are asking what forms of online entertainment and personal services will work for people whose primary access to the network is over their mobile phone, whether in sub-Saharan Africa or on the move in Sussex.
And since they aren't short of ambition they are exploring ways to model the resource consumption of the entire world as a means of understanding the impact of our individual activities and optimising our use of energy and materials.
Siting data centres near wind farms could cut costs
When Prof Hopper gave a speech about the issues to the Royal Society in March most of the coverage focused on his suggestion that we should be siting server farms near to wind farms or other renewable sources of energy.
As he points out, it's a lot cheaper to transmit data than to transmit energy, and the data networks are often already in place, while wind and wave farms in remote sites require extensions to the electricity grid.
It's an eminently sensible idea, but implementing it is a lot more complicated than sending shipping containers full of processors and disks off to remote Scottish islands and congratulating ourselves on the results.
For one thing, we need to make cloud computing work effectively so that processing tasks can be distributed over the network from our desktop computers, laptops and even mobile phones.
At the moment web-based processing is farmed out to Amazon and Google's online services, but when I do a spell check in Word it's stuck on my local processor.
We also need to build data centres that can cope with varying power supplies, so that the server farm uses only the electricity that can be locally generated.
The Cambridge research could be seen as blithe optimism, a desire to avoid facing up to the real implications of a finite world and a growing population, but if there's arrogance here it's the arrogance of the engineer who knows that certain classes of problem are amenable to solution.
I still think of myself as a computer scientist, and I have to point out that I did the Diploma in Computer Science at Cambridge so I'm disposed to look at what they do favourably.
It is not the Earth that needs saving, says Bill Thompson
But the imaginative thinking we're seeing from Prof Hopper and his team, coupled with individual changes to the wasteful way we live here in the West, seem to offer hope for the future of the species.
Of course the Earth will do just fine without us, and we must not confuse working to keep the place habitable for humans with "saving the planet".
As futurist Jamais Cascio pointed out recently on his blog: "the Earth will be just fine. Environmentalism is all about saving ourselves".
If we mess up then it will be hard to tell we were ever here in a few hundred million years or so, apart from the odd radiation hotspot and a layer of soot, plastic and remarkably pure silicon that will be noticed by geologists of whatever species emerges to take our place.
If computing can help defer the day the planet becomes inhospitable to us, I'm all in favour.
Bill Thompson is an independent journalist and regular commentator on the BBC World Service programme Digital Planet.