U.S. supercomputer tops list of world's fastest machines
The U.S. has the world's fastest supercomputer for the first time since 2009, according to a biannual list of Top 500 supercomputers.
The Sequoia is an IBM machine that is part of a generation of IBM supercomputers known as BlueGene/Q. It is based at the U.S. Department of Energy's Lawrence Livermore National Laboratory in California.
The high-performance machine functions at processing speeds of 16.32 petaflops per second (Pflop/s), or about 1.5 million times faster than the average laptop.
One petaflop is equal to one quadrillion, or 10 to the 15th power, floating operations ( i.e. mathematical computations).
The supercomputer is actually a highly interconnected cluster of 1,572,864 processors, or cores, mounted on 98,304 "compute nodes," or boards, that are arranged on a series of 96 standing racks across 318 square metres of floor space.
The biannual list of the world's most powerful machines was revealed Monday at the International Supercomputing Conference in Hamburg.
The list is compiled each June and November by a group of computer experts, manufacturers and computational scientists and uses what's known as the Linpack Benchmark to measure how fast computers execute a particular program.
No. 1 supercomputer used to test nuclear weapons
Supercomputers are used in a variety of fields, including Earth sciences, geophysics, astronomy, medicine and nuclear science.
The newly assembled Sequoia will be used to conduct simulations intended to extend the life of America's aging nuclear weapons arsenal, in lieu of underground nuclear testing.
The supercomputer "will provide a more complete understanding of weapons' performance, notably hydrodynamics and properties of materials at extreme pressures and temperatures," said Thomas D'Agostino of the National Nuclear Security Administration in a news release.
"Computing platforms like Sequoia help the United States keep its nuclear stockpile safe, secure and effective without the need for underground testing," he said.
"While Sequoia may be the fastest, the underlying computing capabilities it provides give us increased confidence in the nation's nuclear deterrent as the weapons' stockpile changes under treaty agreements, a critical part of President Obama's nuclear security agenda."
The U.S. grabbed three of the Top 10 spots on the latest list, two fewer than in November 2011. Its Mira computer at the Argonne National Laboratory in Illinois ranked third, and the Jaguar computer at the Oak Ridge National Laboratory in Tennessee, which was the top U.S. computer on the November 2011 list, in sixth place.
Japan's Fujitsu-made K computer at the RIKEN Advanced Institute for Computational Science, which held the top spot on the previous two lists, was ranked second-fastest, with processing speeds of 10.51 Pflop/s.
Germany's SuperMUC, an IBM machine based at the Rechenzentrum in Leibnitz, came fourth overall but is currently the top supercomputer in Europe. China's Tianhe-1A, developed by its own National University of Defence Technology, rounded out the Top 5.
More than 50% on list commercially owned
Of the 500 computers on the list, more than half are owned by private industry, said Jack Dongarra, a member of the group evaluating the machines and a professor in the innovative computing laboratory of the electrical engineering and computer sciences department of the University of Tennessee.
"Industry gets this — that these computers can provide some strategic advantage, some competitive advantage, and they are willing to purchase those machines," he said.
The list of industries that use supercomputers includes credit card companies, the gaming industry, internet services, financial services, telecommunications, large internet retail companies like Amazon.com and aerospace manufacturers like Lockheed Martin.
The price of the computers near the top of the list runs in the order of hundreds of millions of dollars, Dongarra said.
"The No. 1 machine — I don't know the exact cost, but I would guess it's on the order of $200 million US," he said.
All that computing power also sucks a lot of power. The supercomputers at the top of the rankings consume roughly 10 megawatts of power, or about the equivalent of what 10,000 U.S. households consume in a year, Dongarra said.
"In the U.S., probably If you were to burn one megawatt for one year, that would cost $1 million US. So if our computer system has 10 MW, it would cost $10 million US to run that computer for one year. That's just the power," he said.
The Sequoia is "one of the most energy efficient systems on the list," the group that ranked the computers said in a news release, consuming 7.9 megawatts of power.
The evaluating group said computing speed has evolved considerably in just the six months since the last list was published, with the combined power of all 500 supercomputers on the list totalling 123.4 Pflop/s, compared with 74.2 Pflop/s in November 2011. A total of 170 machines that had made the list in November 2011 failed to make the cut in June 2012.