CHINA NDIYO NCHI PEKEE ILIYOTAJWA KUWA NA COMPUTER YENYE UWEZO MKUBWA KULIKO ZOTE DUNIANI

Massive System in China Named World’s Fastest Computer 

 

A massive supercomputer in China, as expected, has vaulted to the top of a twice-yearly ranking of the world’s fastest computers.
The Tianhe-2 system is a symbol of China’s determination to be a major player at the highest reaches of scientific computation, a discipline historically led by the U.S. that is crucial to advances in fields such as climate change, energy exploration, weapons design and code-breaking for intelligence agencies.
It leapfrogs a system called Titan at Oak Ridge National Laboratory atop the so-called Top500 list being released Monday at a conference in Germany.
Tianhe-2, or Milky Way-2 in English, is powered primarily by the combination of a new model of Intel’s standard Xeon processor used in servers as well as a more specialized chip called the Xeon Phi. But engineers in China also developed adjunct “front-end” chips for the system, the circuitry used to interconnect all the chips as well as its operating system.
In all, the machine has more than three million individual electronic brains, or processor cores, and achieved a speed of 33.86 petaflops–or quadrillion of scientific calculations a second. That’s about 33 million times faster than an iPad, the Top500 organization estimates.
That’s nearly twice as fast as Titan, a Cray system powered by a combination of chips from Advanced Micro DevicesAMD +0.55% and NvidiaNVDA +0.96% that moves to second place in the ranking.
Tianhe-2 is the second machine in China to reach the top of the Top500 list. A predecessor system in 2010 briefly achieved the top speed ranking but was quickly surpassed by machines in the U.S. and Japan.
This time, however, supercomputer experts do not expect U.S. laboratories to field machines that could top the Tianhe-2 until at least 2015.
The Top500 list is compiled by four researchers: 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.