Obama's 2012 Budget Increases Funding for Exascale Supercomputing

With R&D investment shrinking, the U.S. has been losing its leadership position as other nations continue to make rapid progress in science and technology. Recently, China overtook the U.S. in the computing race with the introduction of the fastest supercomputer last October.

The Tianhe-1 machine at the National Center for Supercomputing in the northern port city of Tianjin reported sustained computing of 2.507 Petaflops, significantly faster than the current title holder, the U.S. Department of Energy's (DoE) Cray XT5 Jaguar in Oak Ridge, Tennessee, which is rated at 1.75 Petaflops per second.

To bring U.S. back on track, the Obama Administration's 2012 budget has called for an increase in funding for exascale supercomputing. It is significantly higher than the last budget, which offered only $24 million for supercomputing. The 2012 budget includes $126 million for the development of Exascale supercomputing, according to a report in Wall Street Journal.

The exascale supercomputing systems have the capability to offer processing power that is 1,000 times faster than Petaflop supercomputers.

Under the new budget plan, the DoE’s Office of Science will get $91 million, while the National Nuclear Security Administration will receive $36 million, if the budget is approved by Congress. Also, DoE’s total budget for advanced computing is about $465 million, an increase of 21 percent over 2010, writes WSJ journal.

From homeland security to warnings of global warming, exascale supercomputers will be used to model complex systems of the future. Experts believe that exascale supercomputers will be achievable by 2021.

However, the problem with reaching the exascale milestone is not CPU processing, but its power requirements. According to supercomputing specialist Peter Kogge, the development of exascale systems is liable to hit a power wall, the WSJ journal wrote.

Argonne also envisions Mira as a stepping stone to exascale-class computers that will be faster than petascale-class computers by a factor of a thousand. exascale computing has the potential to address a class of highly complex workloads that have been beyond our reach, not just due to their sheer size, but because of their inherent uncertainties and unpredictability – challenges like understanding the impacts of regional climate change and the design of safe nuclear reactors.