Abstract
Chip multiprocessors (CMPs) can outperform superscalar processors only for the application programs with full of thread-level parallelism. In this paper, we investigate how to combine both the advantages of a CMP and of a superscalar to execute an application. We aggregate multiple processing elements (PEs) in the CMP into a wide superscalar processor. The number of aggregated PEs is dynamically determined by the dependence distance of subsequent thread at the run time. According to different aggregation methods, the new CMP is capable of supporting three execution modes. To obtain the best performance, the processor keeps switching between different execution modes when executing an application, according to the characteristics of subsequent instructions.
| Original language | English |
|---|---|
| Title of host publication | 2005 IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing, PACRIM - Proceedings |
| Pages | 113-116 |
| Number of pages | 4 |
| DOIs | |
| Publication status | Published - 2005 Dec 1 |
| Event | 2005 IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing, PACRIM - Victoria, BC, Canada Duration: 2005 Aug 24 → 2005 Aug 26 |
Publication series
| Name | IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings |
|---|---|
| Volume | 2005 |
Other
| Other | 2005 IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing, PACRIM |
|---|---|
| Country | Canada |
| City | Victoria, BC |
| Period | 05-08-24 → 05-08-26 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Signal Processing
- Computer Networks and Communications
Cite this
}
A dynamically clustered chip multiprocessor. / Wu, Chao Chin; Lai, Guan Joe.
2005 IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing, PACRIM - Proceedings. 2005. p. 113-116 1517238 (IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings; Vol. 2005).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - A dynamically clustered chip multiprocessor
AU - Wu, Chao Chin
AU - Lai, Guan Joe
PY - 2005/12/1
Y1 - 2005/12/1
N2 - Chip multiprocessors (CMPs) can outperform superscalar processors only for the application programs with full of thread-level parallelism. In this paper, we investigate how to combine both the advantages of a CMP and of a superscalar to execute an application. We aggregate multiple processing elements (PEs) in the CMP into a wide superscalar processor. The number of aggregated PEs is dynamically determined by the dependence distance of subsequent thread at the run time. According to different aggregation methods, the new CMP is capable of supporting three execution modes. To obtain the best performance, the processor keeps switching between different execution modes when executing an application, according to the characteristics of subsequent instructions.
AB - Chip multiprocessors (CMPs) can outperform superscalar processors only for the application programs with full of thread-level parallelism. In this paper, we investigate how to combine both the advantages of a CMP and of a superscalar to execute an application. We aggregate multiple processing elements (PEs) in the CMP into a wide superscalar processor. The number of aggregated PEs is dynamically determined by the dependence distance of subsequent thread at the run time. According to different aggregation methods, the new CMP is capable of supporting three execution modes. To obtain the best performance, the processor keeps switching between different execution modes when executing an application, according to the characteristics of subsequent instructions.
UR - http://www.scopus.com/inward/record.url?scp=33746790265&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33746790265&partnerID=8YFLogxK
U2 - 10.1109/PACRIM.2005.1517238
DO - 10.1109/PACRIM.2005.1517238
M3 - Conference contribution
AN - SCOPUS:33746790265
SN - 0780391950
SN - 9780780391956
T3 - IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings
SP - 113
EP - 116
BT - 2005 IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing, PACRIM - Proceedings
ER -