A collaboration from RIKEN, the Tokyo Institute of Technology, and University College Dublin won top place in the June 2014 Graph 500 supercomputer ranking using the K computer, which is located in Kobe, Japan. The results were announced on June 23 at the international conference on high-performance computing (ISC14) being held in Leipzig, Germany.

The Graph 500 ranking is a relatively new benchmark, first issued in 2010, which seeks to gauge the ability of supercomputers on data-intensive loads rather than simple speed, with the goal of improving computing involving complex data problems in five key areas: cybersecurity, medical informatics, data enrichment, social networks, and symbolic networks.

While the TOP500 ranking measures the ability of a computer to solve a system of linear equations with the LINPACK benchmark, in the Graph500 the speed of a breadth-first graph search, measured by number of traversed edges per second (TEPS) is used, with "edges" indicating the connection between two data points. Breadth-first searches, which are commonly used for Big Data applications, involve a substantially larger degree of irregular computations than the LINPACK benchmark.

To conduct the benchmark measurement, Koji Ueno of Tokyo Institute of Technology and RIKEN, along with colleagues, used 65,536 of the K computer’s 88,128 compute nodes, and was able to solve a breadth-first search of an extremely large graph of 1 trillion nodes and 16 trillion edges in 0.98 second. With this achievement it gained the top place with a score of 17,977 gigaTEPS. The K computer was trailed by Sequoia at the Lawrence Livermore Laboratory in the USA, with a score of 16,599 gigaTEPS, and Mira at the Argonne National Laboratory in the USA, with a score of 14,328.

The K computer’s triumph under the new ranking shows that it excels not only at regular parallel computing but also at graph analysis, which relies heavily on irregular computations, and demonstrates the flexibility of K in application to a wide range of applications. According to Kimihiko Hirao, Director of the RIKEN Advanced Institute for Computational Science, "It is also testimony to the talent of the team’s high software development skills which helped them to make optimum use of the K’s powerful hardware."

"We are delighted," he adds, "to have won this prize, as it demonstrates the K computer’s usefulness for tackling complex phenomenon taking place in the real world, which has become increasingly key for computers in recent years."

Two research projects funded by Japan Science and Technology Agency (JST) CREST programs contributed to this achievement: "Advanced Computing and Optimization Infrastructure for Extremely Large-Scale Graphs on Post Peta-Scale Supercomputers" (PI: Prof. Katsuki Fujisawa of Kyushu University and Co-PI: Prof. Toyotaro Suzumura of University College Dublin), which is a project in the research area of Development of System Software Technologies for Post-Peta Scale High Performance Computing (Research Supervisor: Prof. Akinori Yonezawa of RIKEN), and "EBD: Extreme Big Data - Convergence of Big Data and HPC for Yottabyte Processing" (PI: Prof. Satoshi Matsuoka of Tokyo Institute of Technology), which is a project in the Advanced Core Technologies for Big Data Integration area (Research Supervisor: Prof. Masaru Kitsuregawa of National Institute of Informatics).

Contact

Akihiko Okada
Office for Research Communication
RIKEN Advanced Institute for Computational Science

Jens Wilkinson
RIKEN Global Relations and Research Coordination Office
Tel: +81-(0)48-462-1225 / Fax: +81-(0)48-463-3687
Email: pr@riken.jp