Reinventing HPC

TOP500 List at ISC High Performance

TOP500Based on the benchmark known as High Performance Linpack (HPL), the TOP500 list was devised to track the 500 fastest supercomputers in the world. The list was born in 1986 and has been tracking these machines on a continuous basis ever since. Released twice a year – in June and November – the bi-annual TOP500 announcements represent hallmark events for HPC watchers. The next list, the 53rd in its series, will be announced this June at the ISC High Performance conference in Frankfurt, Germany.

As the longest running list of its kind that collects detailed data about these powerful computers, the TOP500 represents a historical record of HPC designs, trends, and technologies. It also serves as a predictor for future HPC performance, spurring development for the fastest supercomputers, while at the same time assisting in planning HPC capacity at supercomputing sites.

The TOP500 has also inspired related rankings, including the Green500, a list that tracks the energy efficiency of the top supercomputers, and the Graph500, which ranks the fastest systems for data-intensive supercomputing.

The most powerful computer on the current list, compiled in November 2018, is the US system Summit, followed by Sierra, both of which are IBM machines. Here is a breakdown of the top 10 supercomputers, along with their HPL rating:

 

  Summit

143,5 petaflops Oak Ridge National Laboratory, United States
  Sierra 94,6 petaflops Lawrence Livermore National Laboratory, United States

  Sunway    TaihuLight

93.0 petaflops National Super Computer Center in Guangzhou, China
  Tianhe-2 61.4 petaflops National Super Computer Center in Guangzhou, China
  Piz Daint

21.2 petaflops 

Swiss National Supercomputing Centre, Switzerland 
  Trinity 20.2 petaflops  Los Alamos National Laboratory, United States
  AI  Bridging   Infrastructure   (ABCI) 19,9 petaflops National Institute of Advanced Industrial Science and Technology (AIST), Japan
  SuperMUC-NG 19,5 petaflops Leibniz Rechenzentrum, Germany
  Titan 17.6 petaflops  Oak Ridge National Laboratory, United States
  Sequoia 17.2 petaflops Lawrence Livermore National Laboratory, United States

 

By registering for the full-conference pass, you will have the opportunity to be a part of the larger community of 3,500 HPC enthusiasts and 160 vendors at ISC High Performance.

On June 17, right before the TOP500 Awarding, Computational scientist and biologist, Prof. Ivo Sbalzarini, will deliver a fascinating conference keynote on how HPC is being used as a tool for scientific investigation and for hypothesis testing, followed by the exhibition opening and the Exhibition Party in the later part of the day.

Register NowRegister for the full-conference pass by May 8 and save 45 percent price reduction off the onsite registration rates.