| Project/Area Number |
25246044
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Computational science
|
|---|
| Research Institution | High Energy Accelerator Research Organization |
|---|
Principal Investigator |
SASAKI TAKASHI 大学共同利用機関法人高エネルギー加速器研究機構, 計算科学センター, 教授 (50259983)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
村上 晃一 大学共同利用機関法人高エネルギー加速器研究機構, 計算科学センター, 講師 (10353369)
藏重 久弥 神戸大学, 先端融合研究環, 教授 (20205181)
山下 智弘 神戸大学, 医学(系)研究科(研究院), 研究員 (20567086)
大町 千尋 名古屋市立大学, 医学(系)研究科(研究院), 研究員 (20588967)
阿蘇 司 富山高等専門学校, 電子情報工学科, 教授 (30290737)
歳藤 利行 名古屋市立大学, 医学(系)研究科(研究院), 研究員 (30377965)
田中 覚 立命館大学, 情報理工学部, 教授 (60251980)
木村 彰徳 足利工業大学, 工学部, 教授 (60373099)
尾崎 正伸 国立研究開発法人宇宙航空研究開発機構, その他部局等, 准教授 (90300699)
尼子 勝哉 大学共同利用機関法人高エネルギー加速器研究機構, 計算科学センター, 研究員 (50044772)
|
|---|
| Research Collaborator |
ASAI Makoto SLAC National Laboratory, SLAC Geant4, プロジェクトリーダー
Gerritsen Margot Stanford University, ICME, 教授
Henderson Nick Stanford University, ICME, 研究員
Incerti Sebastien CERNBG, 研究員
|
|---|
| Project Period (FY) |
2013-10-21 – 2017-03-31
|
| Project Status |
Completed (Fiscal Year 2016)
|
| Budget Amount *help |
¥43,550,000 (Direct Cost: ¥33,500,000、Indirect Cost: ¥10,050,000)
Fiscal Year 2016: ¥10,270,000 (Direct Cost: ¥7,900,000、Indirect Cost: ¥2,370,000)
Fiscal Year 2015: ¥10,270,000 (Direct Cost: ¥7,900,000、Indirect Cost: ¥2,370,000)
Fiscal Year 2014: ¥12,090,000 (Direct Cost: ¥9,300,000、Indirect Cost: ¥2,790,000)
Fiscal Year 2013: ¥10,920,000 (Direct Cost: ¥8,400,000、Indirect Cost: ¥2,520,000)
|
|---|
| Keywords | 放射線シミュレーション / GPGPU / 超並列 / GPU / 並列化 / 放射線 / シミュレーション / 計算物理 / 並列処理 / 並列計算 / マルチスケール |
|---|
| Outline of Final Research Achievements |
Geant4 has been widely used in many fields including physics, medicine, space, engineering and so on. The calculation speed became the key issue to perform more detailed and accurate study. Multi-threaded version of Geant4 was developed for answering such demands. However, the performance was not satisfactory in some areas such as radiotherapy simuation and DNA level simulation. We attmpted further improvements on the performance of MT-Geant4 and evaluated the performance of Xeon Phi in technically and also economilally.
MPEXS was developed based on Geant4 in the CUDA language to be used on NVIDIA GPU. We have achived more than 200 Xeon core equivalent performance on NVIDIA K20c for the case of LINAC treantment simulation. MPEXS-DNA, a CUDA version of Geant4-DNA, has been developed and the performance was almost 250 Xeon core equivalent on K40c. Toward further applications, the development will be contined to include more physics processes such as hadronic interactions.
|
