Three U.S. government scientists are proposing to improve climate change predictions by creating a kind of super-supercomputer.

Michael Wehner and Lenny Oliker of the Lawrence Berkeley National Laboratory and John Shalf of the National Energy Research Scientific Computing Center said one of the greatest challenges to understanding how human activity is changing global climate is not being able to develop accurate cloud simulations.

Although cloud systems have been included in previous climate models, they lack the detail that could improve climate predictions.

Wehner, Oliker and Shalf set out to establish a practical estimate for building a supercomputer capable of creating climate models at 1-kilometer scale -- something not available from existing models, the scientists said. Such a supercomputer would be 1,000 times more powerful than what's available today, but it would cost about $1 billion and require 200 megawatts of electricity.

Their solution is to create such a supercomputer using about 20 million embedded microprocessors. Their computer would cost about $75 million, consume less than 4 megawatts of power and achieve a peak performance of 200 petaflops.

The research appears in the International Journal of High Performance Computing Applications. (c) UPI