NEWS | IN DEPTH
618 9 FEBRUARY 2018 • VOL 359 ISSUE 6376 sciencemag.org SCIENCE
dustry,” Chang says. DOE changed tacks. It
scrapped plans for Aurora, and replaced it
with A21, a machine five times bigger. That
pushed the launch date back to 2021, but
because it was to be the first U.S. exascale
machine, it also effectively pushed up the
U.S. timeline by 2 years.
Skipping the intermediate step of Aurora
is risky, says Kenneth Jansen, an aerospace
engineer at the University of Colorado
in Boulder. “It means one of the stepping
stones is not going to be there.” Still, others say it’s a risk worth taking. “This is the
right way to do it,” says Thom Dunning, a
computational chemist at the University of
Washington in Seattle.
Details of A21’s architecture remain closely
guarded to protect proprietary technology.
But scientists writing software for the new
machine will be given detailed briefings on
the new architecture after they sign nondisclosure agreements. Some of the first
briefings are taking place this week in Knoxville at the second annual Exascale Computing Project meeting.
Researchers already familiar with the
plans say the machine is unlike any they’ve
ever seen before. “A21 is a very different ar-
chitecture,” Chang says. In general terms,
he says, the design focuses on decreasing
the need to move data long distances be-
tween processors, an energetically expen-
sive process. He says the new machine will
likely require 25 to 30 megawatts of power,
only about twice that of Summit. Asked
whether he thinks Intel will be able to pull
off the new architecture, Chang says, “I am
confident they will.”
One outside challenge could be money.
Congress has yet to pass the fiscal 2018
budget, and instead has funded the government through a series of continuing resolutions that keep funding levels the same as
the prior year while forbidding the launch
of new projects, such as building the A21
machine. For now, that’s not a problem, be-
cause DOE is still able to support the under-
lying scientific developments as part of its
existing Exascale Computing Project, says
Jack Dongarra, a supercomputing expert
at the University of Tennessee in Knoxville.
But soon it will be time to start fabricating
chips for A21, which is expected to cost between $300 million and $600 million, according to market research firm Hyperion
Research. “In 2021 will the budget be there
to do this?” asks Horst Simon, a supercomputing expert and deputy director of
the Lawrence Berkeley National Laboratory
in Berkeley, California. “I don’t know.” j
China’s planned exascale computer threatens Summit’s position at the top
K computer, 10.0
Sunway TaihuLight, 93
Since 2013, China has operated the most powerful supercomputer in the world. Summit is likely to reclaim the
title for the United States this year. China is on track to unveil the first exascale computer in 2020.
This summer, when engineers flip the switch on Summit, a supercomputer being assembled at Oak Ridge National Lab- oratory in Tennessee, the machine is expected to be the most powerful in the world. That would return the United States to the top of the supercomputing rankings for the first time since
June 2013, when it lost the top spot to Tianhe-2, a machine housed
at China’s National Supercomputer Center in Guangzhou.
“Of course, we hoped we could have [the top machine] for a
longer time,” says Depei Qian, a computer scientist at Beihang
University in Beijing. “But the government agencies understand
that no country can be No. 1 forever.”
In the global game of supercomputing leapfrog, China is likely
to take back the title from the United States when it builds the
first exascale computer: a machine capable of 1 billion billion
floating-point operations per second, or 1 exaflop. Under the
country’s 13th 5-Year Plan, released in 2015, China is committed
to launching its first exascale supercomputer by the end of 2020.
That could put it a full year ahead of A21, the first U.S. exascale
supercomputer, planned for launch in 2021 (see main story,
p. 617). Japan is also aiming for an exascale machine with a suc-
cessor to its K supercomputer at the RIKEN Advanced Institute
for Computational Science in Kobe, though the head of the proj-
ect has said that its delivery could slip to 2021 or 2022. Finally,
the European Union could cross the exascale threshold in 2021,
according to market research firm Hyperion Research.
But that won’t be the end of the race. The four supercomputing
powers are convinced they need to push the frontiers in order to
compete in a wide range of scientific disciplines, defense technology, industrial technology, and computer products. “Everybody
is moving as fast as they can,” says Jack Dongarra, a supercomputing expert at the University of Tennessee in Knoxville who
keeps close tabs on international supercomputing efforts. And
once they cross one threshold, he says, “then it’s on to the next
one.” —Robert F. Service
With additional reporting by Dennis Normile.