China has replaced United States as the maker of the world's fastest supercomputer.
A Chinese research center unveiled the Tianhe-1A at a national conference on high-performance computers in China.Tianhe, meaning Milky Way, was designed by the National University of Defense Technology (NUDT) in China, usurped the Cray XT5 Jaguar system as the world's fastest supercomputer.
The latest supercomputer, which is the result of two years of work by 200 scientists, is housed at National Supercomputer Center in Tianjin and is already fully operational. Made at a cost of over $88 million, It has set a new performance record of 2.507 petaflops, or quatrillion calculations per second, making it the fastest system in the world today. Ordinarily the computer runs at 563.1 teraflops, or trillion calculations per second.
In practical terms, the computer will run at about 1.2 petaflops. As of June 2010, the 500 fastest supercomputers in the world combine for 32.4 petaflops of computing power.
NUDT president Zhang Yulin said the 155-ton system, with 103 refrigerator-like cabinets lined up on an area of about 1,000 square meters, is expected to process seismic data for oil exploration, conduct bio-medical computing and help design aerospace vehicles.
Tianhe-1A is powered by 7,168 Nvidia Tesla M2050 graphics processor units (GPUs) and 14,336 Intel Xeon central processing units (CPUs). It would require more than 50,000 CPUs and twice as much floor space to deliver the same performance using CPUs alone.
the Cray XT5 Jaguar system has more than 19000 computers and 224,000 processing elements, based on standard AMD processors.
The performance and efficiency of Tianhe-1A was simply not possible without GPUs, said Guangming Liu, chief of National Supercomputer Center in Tianjin. The scientific research that is now possible with a system of this scale is almost without limits; we could not be more pleased with the results.
GPUs are redefining high performance computing, said Jen-Hsun Huang, president and CEO of Nvidia. With the Tianhe-1A, GPUs now power two of the top three fastest computers in the world today. These GPU supercomputers are essential tools for scientists looking to turbocharge their rate of discovery.
Tianhe-1A is also energy efficient, saving enough power to provide electricity to over 5,000 homes for a year, according to a statement from Nvidia. Tianhe-1A consumes only 4.04 megawatts, making it 3 times more power efficient compared to a 2.507 petaflop system built entirely with CPUs that would consume more than 12 megawatts. However, its annual electricity bill could still touch $2.7 million.
As programmers and big semiconductor firms have focused traditionally on CPUs, graphic chips have been underutilized and analysts say that they should be tapped to improve the performance of everyday computers ranging from desktops to tablets.
The battle to build the fastest supercomputer has become a point of national pride as these high performance computers are used in a areas like defense, energy, finance, science. They are also used for drug discovery, hurricane and tsunami modeling, cancer research, car design, and studying the formation of galaxies.
For example, oil companies use supercomputers to find reservoirs, while Wall Street traders use it for speedy automated trades.
The United States has seven of the world's most powerful computers, while China has two and Germany has one, according to Nvidia.
Yulin said the technical data of Tianhe-1A had been submitted to the world Top 500 list, compiled by the University of Mannheim, in Germany, the National Energy Research Scientific Computing Center, the Lawrence Berkeley National Laboratory, and the University of Tennessee in the United States.
The next Top 500 supercomputer list will be released in November.
Of the world's fastest 500 supercomputers, the United States has 291, including the top 10, while Europe has 145 and Asia 49, according to the latest list.
Meanwhile, China's Dawning Information Industry Company is also working on a supercomputer that could overcome the petaflop barrier.