WASHINGTON - President Barack Obama's federal budget, released Wednesday, calls for increases in federal research and development spending. The overall spending increase, though shows a decline when adjusted for inflation.
The White House has proposed a $3.8 trillion budget for the 2014 fiscal year, which begins Oct. 1.
This budget plan would set aside $142.8 billion for federal R&D programs, an increase of $1.9 billion. or 1.3%, from fiscal 2012. The current U.S. fiscal year, 2013, is being funded under a series of continuing resolutions because a budget was never formally adopted.
With inflation between 2012 and 2014 expected to be in the range of 4.0%, "we have to admit a decline in real terms over this period," said John Holdren, director of the White House Office of Science and Technology Policy, at a briefing today. "But it's a small decline and it represents a continuing strong commitment from the President and from this administration,"
China is expected to exceed the U.S. in R&D spending in about 10 years, according to Battelle, a research and technology development company.
Federal R&D operations cover a wide range of scientific interests, including health, cybersecurity and supercomputing.
China and Europe are investing heavily in building supercomputers -- China is expected to deliver two 100 petaflop machines in 2015, said Earl Joseph, an analyst at IDC.
A petaflop is 1,000 trillion (one quadrillion) sustained floating point operations per second, and the next big leap, exascale is 1,000 petaflops.
"I think China is definitely going to leapfrog us in 2015," said Joseph, who said he wouldn't be surprised if China delivered a 100 petaflop system by the end of 2014.
The most powerful system today, called Titan, is a Cray system installed at Oak Ridge National Lab and is capable of nearly 18 petaflops.
Holdren said the U.S. Dept. of Energy's budget plan calls for increased spending on high performance computing research, but "does not set a goal for building exascale" systems.
China's push is important because it is learning how to design high-end machines from the processor level on up, said Joseph. "Five to 10 years down the road, they could take leadership in that," he added.
U.S. work on high performance computing technology has been hurt by budget disputes between Congress and the White House in recent years.
Every decade, compute power has double in alignment with Moore's Law. The power of supercomputing has followed a similar path. Following that target, an exascale system might have been possible in the 2018 timeframe. The petaflop barrier was broken in 2008.
Exascale development, however, poses unique challenges because of the enormous amount power such a system would likely need.
Sign up for Computerworld eNewsletters.