The US can once again boast the world’s most powerful supercomputer, after two and a half years during which Japan has held the crown.
The IBM BlueGene/Q system, named Sequoia, is installed at the Department of Energy’s Lawrence Livermore National Laboratory and hit 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores. It’s also one of the most energy efficient systems on the list.
Sequoia will from next year used to carry out simulations for nuclear weapons, eliminating the need for expensive and postentially dangerous underground tests.
Fujitsu’s K Computer, at the RIKEN Advanced Institute for Computational Science (AICS) in Kobe, Japan, now slides to second place with 10.51 Pflop/s using 705,024 SPARC64 processing cores. It’s held the number one spot for the last year.
The new Mira supercomputer, an IBM BlueGene/Q system at Argonne National Laboratory in Illinois, came in at number three, with 8.15 petaflop/s. The upgraded Jaguar at Oak Ridge National Laboratory in Tennessee now clocks in at number six.
For the first time in a while, European systems make a good showing. Fourth on the list is SuperMUC, an IBM iDataplex system installed at Leibniz Rechenzentrum in Germany. Another German machine, the JuQUEEN BlueGene/Q at Forschungszentrum Juelich, is number eight, pipped by Italy’s IBM BlueGene/Q system, installed at CINECA.
China, which briefly took the first and third spots in November 2010, has two systems in the Top 10, with Tianhe-1Aat the National Supercomputing Center in Tianjin at number five and Nebulae at the National Supercomputing Centre in tenth position.
IBM’s still the leading vendor, boasting four of the top 10 spots. It now has 213 systems in the top 500, compared to HP with 138.
Performance generally is improving fast.
“Total performance of all the systems on the list has increased considerably since November 2011, reaching 123.4 Pflop/s. The combined performance of the last list was 74.2 Pflop/s,” say the compilers of the Top500 list.
“In all, 20 of the supercomputers on the newest list reached performance levels of 1 Pflop/s or more. The No. 500 machine on the list notched a performance level of 60.8 teraflop/s, which was enough to reach No. 332 just seven months ago.”
Intel’s share of the market has declined a little, from 76.8 percent six months ago to 74.4 percent now. The AMD Opteron family trails at quite some distance, with 12.6 percent.
Now, 58 systems use accelerators or co-processors,up from 39 six months ago. Of these, 53 use NVIDIA chips, two use Cell processors, two use ATI Radeon and one new system uses Intel MIC technology.
The list is published every six months by Hans Meuer of the University of Mannheim, Germany; Erich Strohmaier and Horst Simon of Lawrence Berkeley National Laboratory; and Jack Dongarra of the University of Tennessee, Knoxville.