We got an insight into the TOP500 update, the list of the world’s fastest supercomputers, with the Department of Energy’s Oak Ridge National Laboratory leading the list. When it debuted earlier this month, Summit managed to clock in at 122 petaflops of compute power when exposed to High Performance Linpack (HPL) which is a benchmark used by scientists to make a ranking list of supercomputers in the TOP500 list.
Summit is equipped with 27,000 Nvidia graphics processing unit chips (GPU). It’s worth adding that five out of seven of the fastest supercomputers in the world sport Nvidia GPUs, including the Tesla V100, originally built in May 2017. Scientists already have used Summit in different science branches and applied machine learning algorithms in order to search genetic association between different diseases, as well as learn more about compounds of superconductors.
“When we first started talking about the original Tesla K80 back in 2015, we were only contributing about 11 percent of the list that year, if I add up all the computational horsepower on the top of the list,” Nvidia VP Ian Buck told VentureBeat. “This year, the majority of 56 percent of the computation on the list is coming from GPUs, and this really talks to the adoption of accelerated computing, of using GPUs for solving the kinds of problems and building the kinds of systems that are necessary to advance computing.”
Another addition to the list is Sierra, that sits at the DOE Lawrence Livermore National Laboratory. At the moment, Sierra ranks as the third fastest supercomputer, and is equipped with 71 petaflops of compute power. Both mentioned supercomputers were designed by IBM. Furthermore, they use IBM Power9 CPU.
According to the report in VentureBeat, the new rankings were announced today at the International Supercomputing Conference that is taking place this week in Frankfurt, Germany. The report further writes that Nvidia debuted nine new GPU Cloud computing containers that would make it easier for users to use deep learning frameworks.
Back at the top
The U.S. came back to the top of the list by owning the world’s fastest supercomputer after China was dominant in this field for years. China’s supercomputer Sunway TaihuLight is equipped with 93 petaflops of compute power that receives power from the Sunway SW26010 CPU. The supercomputer is housed at the National Supercomputing Center in Wuxi and it was dominant at the TOP500 list for the last two years.
Another supercomputer, Titan, holds the seventh place of the world’s fastest supercomputer. It is based at the Oak Ridge National Laboratory and also utilizes Nvidia GPU. Back in 2012, Titan was rated as the world’s fastest supercomputer in the November 2012 TOP500 edition, although from 2013 until now, China was at the very top, with Tianhe-2, the Chinese supercomputer that was built by China’s National University of Defense Technology.
Many of the supercomputers in the world are being used by governments. They use the supercomputers in order to perform high demanding tasks such AI model training and are handling massive amounts of data. The world’s fifth fastest supercomputer, which is also considered the fastest in Japan, is called Artificial Intelligence Bridging Cloud (AIBC) because it is training AI models and working with a lot of data.
