High Performance in Big Data Analysis
The K computer took first place in the Graph500 rankings announced in June 2016*. This marks the third consecutive time following November 2015 (four times in total).
In the Top500 rankings, which evaluate supercomputer computational performance, the taking of first place by Sunway TaihuLight from China attracted attention. However, in the Graph500 rankings, which evaluate large-scale graph analysis performance, the K computer continues to hold first place, maintaining a large lead over the other machines. This demonstrates that the K computer has particularly high performance in the field of big data analysis.
* Analysis results by an international collaborative research group consisting of Kyushu University, the Tokyo Institute of Technology, RIKEN, the Barcelona Supercomputing Center in Spain, and Fujitsu.
** A numerical value representing the scale of a problem calculated in the Graph500 benchmark.
Fujitsu Challenges Itself to Take on Projects in the Fields of Space, Meteorology, Astronomy, and so on to Make Human Dreams Come True
The term "graph" in Graph500 refers to the relationships between data sets; analyzing these relationships at high speed is called "graph analysis." Thus far, supercomputers have mainly been used for numerical calculations such as simulations. Recently, however, since big data analysis has become increasingly important, "graph analysis" performance is attracting more attention.
Fujitsu developed the "K computer" supercomputer in collaboration with RIKEN. The K computer was completed in June 2012, and it was opened up for shared use by researchers and engineers in September 2012 to help them solve various problems.
Fujitsu will continue to challenge itself to tackle various projects to make human dreams come true in the fields of science and technology, including space, meteorology, astronomy, and advanced research.