抄録
The very nature of distributed-memory parallel architectures demands the serious consideration of interprocessor communications, by which non-local memory access can be implemented. In this paper, we propose a method for optimization that utilizes asynchronous and bulk communications. We constructed an HPF- compiler, a subset of HPF, and evaluated it on the CM5, a true distributed-memory machine using three non-trivial benchmark programs. With optimized code, there was considerable improvement in communications over non-optimized code and we had much better results than when optimizing by means of CM Fortran.
| 元の言語 | 英語 |
|---|---|
| ページ | 185-189 |
| ページ数 | 5 |
| 出版物ステータス | 出版済み - 1 1 1995 |
| イベント | Proceedings of the 1995 Conference on Supercomputing - Barcelona, Spain 継続期間: 7 3 1995 → 7 7 1995 |
その他
| その他 | Proceedings of the 1995 Conference on Supercomputing |
|---|---|
| 市 | Barcelona, Spain |
| 期間 | 7/3/95 → 7/7/95 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Computer Science(all)
これを引用
Using asynchronous and bulk communications to construct an optimizing compiler for distributed-memory machines with consideration given to communications costs. / Sato, Hiroyuki; Nanri, Takeshi; Shimasaki, Masaaki.
1995. 185-189 論文発表場所 Proceedings of the 1995 Conference on Supercomputing, Barcelona, Spain, .研究成果: 会議への寄与タイプ › 論文
}
TY - CONF
T1 - Using asynchronous and bulk communications to construct an optimizing compiler for distributed-memory machines with consideration given to communications costs
AU - Sato, Hiroyuki
AU - Nanri, Takeshi
AU - Shimasaki, Masaaki
PY - 1995/1/1
Y1 - 1995/1/1
N2 - The very nature of distributed-memory parallel architectures demands the serious consideration of interprocessor communications, by which non-local memory access can be implemented. In this paper, we propose a method for optimization that utilizes asynchronous and bulk communications. We constructed an HPF- compiler, a subset of HPF, and evaluated it on the CM5, a true distributed-memory machine using three non-trivial benchmark programs. With optimized code, there was considerable improvement in communications over non-optimized code and we had much better results than when optimizing by means of CM Fortran.
AB - The very nature of distributed-memory parallel architectures demands the serious consideration of interprocessor communications, by which non-local memory access can be implemented. In this paper, we propose a method for optimization that utilizes asynchronous and bulk communications. We constructed an HPF- compiler, a subset of HPF, and evaluated it on the CM5, a true distributed-memory machine using three non-trivial benchmark programs. With optimized code, there was considerable improvement in communications over non-optimized code and we had much better results than when optimizing by means of CM Fortran.
UR - http://www.scopus.com/inward/record.url?scp=0029194713&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0029194713&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:0029194713
SP - 185
EP - 189
ER -