Abstract
This paper introduces a new mechanism for the exposure of large grain parallelism. The scheme performs lazy task creation; inlining all tasks provisionally and extracting parallelism from the inlined information later on demand. However, unlike other mechanisms, the further task demand is satisfied by the next evaluation stream rather than retrospectively reversing the inlining decision of the current stream. The scheme is called lazy decomposition because decomposition itself is throttled rather than just the extraction of a task. Lazy decomposition makes the serial section clearly separated from the parallel section in an evaluation tree for a particular function, and this allows the serial section to adopt a sequential algorithm. The performance improvement is significant in divide-and-conquer applications by adoption of sequential algorithms.
| Original language | English |
|---|---|
| Title of host publication | 1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997 |
| Editors | Wanlei Zhou, Andrzej Goscinski, Michael Hobbs |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 427-434 |
| Number of pages | 8 |
| ISBN (Electronic) | 0780342291, 9780780342293 |
| DOIs | |
| Publication status | Published - 1997 Jan 1 |
| Event | 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997 - Melbourne, Australia Duration: 1997 Dec 10 → 1997 Dec 12 |
Publication series
| Name | 1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997 |
|---|
Other
| Other | 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997 |
|---|---|
| Country | Australia |
| City | Melbourne |
| Period | 97/12/10 → 97/12/12 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications
- Hardware and Architecture
- Signal Processing
Cite this
}
Lazy decomposition : A novel technique to control parallel task granularity. / Hwang, Suntae; Cha, Hojung.
1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997. ed. / Wanlei Zhou; Andrzej Goscinski; Michael Hobbs. Institute of Electrical and Electronics Engineers Inc., 1997. p. 427-434 651511 (1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Lazy decomposition
T2 - A novel technique to control parallel task granularity
AU - Hwang, Suntae
AU - Cha, Hojung
PY - 1997/1/1
Y1 - 1997/1/1
N2 - This paper introduces a new mechanism for the exposure of large grain parallelism. The scheme performs lazy task creation; inlining all tasks provisionally and extracting parallelism from the inlined information later on demand. However, unlike other mechanisms, the further task demand is satisfied by the next evaluation stream rather than retrospectively reversing the inlining decision of the current stream. The scheme is called lazy decomposition because decomposition itself is throttled rather than just the extraction of a task. Lazy decomposition makes the serial section clearly separated from the parallel section in an evaluation tree for a particular function, and this allows the serial section to adopt a sequential algorithm. The performance improvement is significant in divide-and-conquer applications by adoption of sequential algorithms.
AB - This paper introduces a new mechanism for the exposure of large grain parallelism. The scheme performs lazy task creation; inlining all tasks provisionally and extracting parallelism from the inlined information later on demand. However, unlike other mechanisms, the further task demand is satisfied by the next evaluation stream rather than retrospectively reversing the inlining decision of the current stream. The scheme is called lazy decomposition because decomposition itself is throttled rather than just the extraction of a task. Lazy decomposition makes the serial section clearly separated from the parallel section in an evaluation tree for a particular function, and this allows the serial section to adopt a sequential algorithm. The performance improvement is significant in divide-and-conquer applications by adoption of sequential algorithms.
UR - http://www.scopus.com/inward/record.url?scp=85060735816&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85060735816&partnerID=8YFLogxK
U2 - 10.1109/ICAPP.1997.651511
DO - 10.1109/ICAPP.1997.651511
M3 - Conference contribution
AN - SCOPUS:85060735816
T3 - 1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997
SP - 427
EP - 434
BT - 1997 3rd International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 1997
A2 - Zhou, Wanlei
A2 - Goscinski, Andrzej
A2 - Hobbs, Michael
PB - Institute of Electrical and Electronics Engineers Inc.
ER -