21 July 2014

From Pel
Jump to navigation Jump to search

Prof. Toyotaro Suzumura and his team win first place in Graph 500 “Big Data” supercomputer ranking!

Prof. Toyotaro Suzumura, a Visiting Associate Professor in the Performance Engineering Lab at the School of Computer Science and Informatics, and his JST CREST team have won first place in the latest Graph 500 “Big Data” benchmarking ranking. His team used the K computer, located in Kobe, Japan, in collaboration with RIKEN and Tokyo Institute of Technology. The results were announced on June 23 at the International Supercomputing Conference (ISC14) held in Leipzig, Germany.

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

By carrying out a series of optimizations at the level of algorithms and software implementation, Prof. Suzumura’s team was able to solve a “breadth-first search” of an extremely large graph (over 1 billion nodes and 17 billion edges) in less than 1 second. With this achievement it placed first in this year’s Graph 500 ranking. In second place was the Sequoia computer at Lawrence Livermore Laboratory in the USA; the Mira computer at the Argonne National Laboratory in the USA was in third place.