Where does all the computing power go? A scientific example.

Source: The Economist

NEON, the National Ecological Observatory Network, is being created with funding from the U.S. National Science Foundation (using a Congress to earmark $434 million). The operating budget will be around $80m a year. The project is to monitor multiple sites in each zone with complex instrumentation for 30 years. The map above shows the way that the United States has been divided into ecological zones for the study.

I quote from the article in The Economist describing the study:

Crucially, these instruments will take the same measurements in the same way in every place. By gathering data in this standardised way, and doing so in many places and over long periods of time, Dr Schimel hopes to achieve the statistical power needed to turn ecology from a craft into an industrial-scale enterprise. The idea is to see how ecosystems respond to changes in climate and land use, and to the arrival of new species. That will let the team develop models which can forecast the future of an ecosystem and allow policymakers to assess the likely consequences of various courses of action....... 

When fully operational NEON is expected to generate 200 terabytes a year. That is four times as much as the Hubble space telescope, a reasonably big piece of science, churned out in its first two decades.

Of course, this implies huge data storage capacity and information processing capacity, not to mention the software necessary to reduce the data to representations that the scientists can interpret and study. I suspect that few countries could mount similar studies since few would have the money, the scientists, or the ICT capacity.