Study of highly scalable parallel I/O systems for high performance computing with huge data management
| Project/Area Number |
25330148
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
High performance computing
|
|---|
| Research Institution | Institute of Physical and Chemical Research |
|---|
Principal Investigator |
Tsujita Yuichi 国立研究開発法人理化学研究所, 計算科学研究機構, 開発研究員 (70360435)
|
|---|
| Project Period (FY) |
2013-04-01 – 2017-03-31
|
| Project Status |
Completed (Fiscal Year 2016)
|
| Budget Amount *help |
¥4,940,000 (Direct Cost: ¥3,800,000、Indirect Cost: ¥1,140,000)
Fiscal Year 2015: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2014: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2013: ¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
|
|---|
| Keywords | ネットワークコンピューティング / 並列入出力 / PCクラスタ / MPI / 並列ファイルシステム / MPI-IO / アグリゲータ / 並列I/O / マルチスレッド処理 / CPUコア配置 |
|---|
| Outline of Final Research Achievements |
ROMIO is well known as one of the representative MPI-IO implementations, and it provides high performance parallel I/O. However, it can not achieve enough I/O performance when we have multiple processes per node. In this research work, we have introduced topology-aware treatment in layout of processes that play file I/O and data aggregation task associated with the file I/O task in order to improve I/O performance of ROMIO.
Consequently, we have achieved high scalable parallel I/O in ROMIO using the proposed scheme with any kind of process layout.
|
Report
(5 results)
Research Products
(5 results)
