This category is a successor of the "Parallel Environment" of the last two contests. Although the method of exploiting parallelism was one of the factors that was judged in the Parallel Environment, speed is top priority above all others in this category.
The speed category had eight submissions in total. While five of them worked for eight large-scale test data, three did not.
One of the five won first prize. The program scored top marks for almost all test data. Although execution using fewer processors achieved better results in the initial stage, the use of more processors caught up with that using fewer processors in the middle, and scored top marks in the final.
Four other programs won second prizes, because each had one or more of the following advantages:
Although the remaining three programs did not work for the large-scale test data because of memory or time limitations, they did work for smaller-scale test data. Since one of the three was designed by the participant who won second prize, the remaining two programs were awarded honorable mentions.
Generally speaking, the submitted programs were well designed. They all worked for small-scale test data, and most of them successfully exploited parallelism.
On the other hand, all programs adopted conventional search methods, and no novel methods could not be found. The problem might thus have been too difficult for testing new ideas.