In India, one reason I see is the [traditional power] grid might not be able to keep up with the data centre growth. Indeed, if you go to India you might see that a lot of things are on diesel generators. Running a data centre on diesel generators is not very cost effective. In order to build cost effective IT services for billions like the autoriksha driver in India, we need to build a data centre that uses energy very efficiently. So I see India as a big opportunity because undoubtedly IT services will grow, and India has a history of indigenous development in this areaglobal gas as it is called in India. This is well known, so I expect that this is the place where the opportunity exists. The only problem in India is the farms are not very large. So what is the opportunity? Well, in India there are big cooperatives that are very successful in collecting milk and taking it to the dairy like the Gujarat co-operative Milk Marketing Federation (GCMMF), which is behind the Amul brand.
Amul is very successful. Someone who can collect milk and take it to the dairy can do business under the GCMMFs model. An ordinary person can sell even just a cup of milk under this co-operative setup.
Is it possible to do the same thing with manure? We believe so, though we havent done a detailed study. We think we can pool the manure and have a data centre partially powered by manure and biogas.
You mentioned India and China. Those particular geographies may not have the legacy issues of perhaps the U.S., where there are huge cities feeding off existing dirty power grids. Would you say that in developed economies the choice of going with dirty power will win out simply because the old power grids are more cost effective?
Yes, you are right. If I understand your question right, electricity may be a lot cheaper in America in various locations so people may just go with that instead of doing this. Is that correct?
We believe that trend will change.
The original trend we saw was a lot of people actually going to locations where electricity was cheap. That was the first wave. So as we moved to these plants, we got electricity at a lower cost. But with such a setup, at the end of the line somewhere somebody had to pay more. We believe that while that was the trend in the beginning, it will change because, economically in the US as well, it will be important to colocate and have your own source of power.
Now if you look at India, electricity is not cheap. In fact, when we help build data centres in India, the grid can only supply us a fraction of the amount or none at all, and we have to use diesel generators. And when you use diesel, the electricity cost is almost US$0.25 per Kilowatt hour. In the US, the cost might be US$0.10 cents per Kilowatt hour. In India, if the data centres were to grow, it will be commensurate with the national GDP [gross domestic product], the economic growth. If the data centres were to scale up, alternatives to diesel would have to be seriously looked at. The payback period on what were proposing is in the order of two years.
Sign up for Computerworld eNewsletters.