Image via Wikipedia
First some facts about the current state of energy consumption in the world of Cloud Computing.
Gartner’s Data Center conference offers these facts:
In an article about Cloud Computing and Environment, The Economist points out to the following dark fact. (Emphasis is mine).
They are huge energy hogs: in America alone, according to the country's Environmental Protection Agency (EPA), data centres already account for 1.5% of electricity consumption.
And this figure is growing. Data centres consumed 0.6% of the world's electricity in 2000, and 1% in 2005. Globally, they are already responsible for more carbon-dioxide emissions per year than Argentina or the Netherlands, according to a recent study by McKinsey, a consultancy, and the Uptime Institute, a think-tank. If today's trends hold, these emissions will have grown four-fold by 2020, reaching 670m tonnes. By some estimates, the carbon footprint of cloud computing will then be larger than that of aviation.
In fact, these datacenters consume more power than some of the world’s fastest supercomputers.
The problem then will be building the computer without needing a nuclear reactor to power it. IBM's Roadrunner, the fastest supercomputer in the world, used to simulate mushroom clouds by Los Alamos National Labs in New Mexico, was the first to break the petaflop barrier, at 1 trillion computations per second. It eats 2.5 megawatts of power. And that's nothing compared with the giant data centers being built around the world by the likes of IBM, Hewlett-Packard, Google, Yahoo! and Microsoft. The biggest of these consumes 250 megawatts.
What we are facing here is not a simple problem which we can shrug off as collateral damage. It is a problem that could drag down the Cloud Computing industry like the current day aviation industry. If it is ignored, the big three of cloud computing could face the same fate faced by the big three in Detroit today. The only way to avoid any government regulations on energy usage in data centers is by self regulating the industry on our own. Sorry for the rant-like exaggeration but the situation is serious and we need fast solutions.
Let us explore the possible advantages of pursuing Green Technologies. First, any energy savings in the datacenter operations means less Op-ex for enterprises and SaaS providers. Second, less Op-ex for enterprises and SaaS providers, in turn, means further cost savings for Cloud Computing customers who are already enjoying considerable savings in Cloud Computing due to factors like multi-tenancy, etc.. Finally, any attempt to extend the life of the planet for human habitation means that the fruits of our entrepreneurship can live well beyond our own life. If you also consider the possibility of mushroom like growth of a new business field based entirely on Green Technologies, we have a win-win reason to spend our mind and resources on fixing the above said energy problem.
I do agree that companies like IBM, Microsoft, Google, etc., are spending money on lowering the energy costs. Google is trying to float their datacenters on the ocean waters, Intel is trying to see if they can use outside air for datacenter cooling, Microsoft is trying to setup an economical datacenter inside a tent, Iceland is trying to lure different companies to invest in datacenters there (Also, read this good Business Week article on the topic), etc. All these attempts are pretty good and some of them will definitely lead to significant breakthroughs. But, there is a real danger of such breakthrough research getting locked up within a smaller number of companies due to their patents. Plus, these patented results, while resulting in Greener Technologies, will not offer the cost savings, expected with the adaption of such technologies, to other cloud vendors.
There is an opportunity for us to get both the Green Technologies and the associated cost savings. We could tap into the vast expertise of academic community whose hands are tied due to drying up of research grants. If Cloud Computing vendors, both the giants as well as the mom and pop kind, come together and start a Global Green Computing Fund, we can, then, fund various research projects in the laboratories of our academic institutions, tapping some of the best minds available there. This could result in newer technologies for reducing the energy costs for all the cloud computing vendors instead of just the elite few. It is time for evangelists and vendors to think about such a fund and how it can benefit Cloud Computing as a whole. Companies like Google and Microsoft can also benefit from this fund by diverting just a portion of what they spend on Green Computing research.
Well, this might sound more like an overambitious effort for social good but I, mainly, see the cost savings for Cloud Computing vendors with very little investment from their side. I am not a global player who could kick start such an effort. I am just throwing out this idea in open to see if it picks up steam, at least, with the pundits and evangelists of Cloud Computing. Feel free to add your comments on the viability of this idea.
Post Comment