| Project/Area Number |
17K00171
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
High performance computing
|
|---|
| Research Institution | Osaka Metropolitan University (2022)
Osaka Prefecture University (2017-2021) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-04-01 – 2023-03-31
|
| Project Status |
Completed (Fiscal Year 2022)
|
| Budget Amount *help |
¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000)
Fiscal Year 2019: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2018: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
Fiscal Year 2017: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
|
|---|
| Keywords | 高性能計算 / 組み合わせ最適化 / アルゴリズム / ハイパフォーマンス・コンピューティング / 応用数学 |
|---|
| Outline of Final Research Achievements |
We have developed a parallel implementation of a new method (sequential algorithm proposed by the applicant) for obtaining the optimal solution in pseudo-polynomial time and polynomial space for the subset sum problem of combinatorial optimization problems, using multi-core CPUs, whose performance has been improving with the increase in the number of cores, and GPUs, which are currently attracting attention and are promising parallel computing platforms for the future. In addtion, to facilitate the development of parallel implementations using multiple GPUs, we have also developed a prototype implementation that performs multiple-precision multiplication by reducing it to a matrix product. By the reduction, it is possible to execute multiple-precision multiplication in parallel at high speed using a parallel implementation of the matrix product (for multiple GPUs) provided by a GPU vendor, etc.
|
| Academic Significance and Societal Importance of the Research Achievements |
研究代表者が2016年12月に発表した，NP困難な組み合わせ最適化問題の部分和問題とナップサック問題を対象に最適解を擬似多項式時間および多項式空間で厳密に求める新しい手法（逐次アルゴリズム）を，年々コア数の増加により性能が向上しているマルチコアCPUや，現在注目されており将来も有望な並列計算プラットフォームであるGPUを用いた並列処理により高速化することにより，部分和問題を現実的なメモリ消費量と許容できる計算時間で解けるようにしたこと．
|
