Distributed Models of Thread Level Speculation.
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Distributed Models of Thread Level Speculation. / Oancea, Cosmin Eugen; Selby, Jason W. A.; Giesbrecht, Mark W.; Watt, Stephen M.
Proceedings of International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA'05). Vol. 5 CSREA Press, 2005. p. 920-927.Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Distributed Models of Thread Level Speculation.
AU - Oancea, Cosmin Eugen
AU - Selby, Jason W. A.
AU - Giesbrecht, Mark W.
AU - Watt, Stephen M.
N1 - @inproceedings{oancea2005distributed, title={Distributed Models of Thread Level Speculation.}, author={Oancea, Cosmin E and Selby, Jason WA and Giesbrecht, Mark and Watt, Stephen M}, booktitle = {Proceedings of International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA'05)}, pages = {920-927}, year = {2005}, publisher = {CSREA Press} }
PY - 2005
Y1 - 2005
N2 - This paper introduces a novel application of thread-level speculation to a distributed heterogeneous environment. We propose and evaluate two speculative models which attempt to reduce some of the method call overhead associated with distributed objects. Thread-level speculation exploits parallelism in code which is not provable free of data dependencies. Our evaluation of applying thread-level speculation to client-server applications resulted in substantial performance increases, on the order of 3 times for ourinitial model, and 21 times for the second.
AB - This paper introduces a novel application of thread-level speculation to a distributed heterogeneous environment. We propose and evaluate two speculative models which attempt to reduce some of the method call overhead associated with distributed objects. Thread-level speculation exploits parallelism in code which is not provable free of data dependencies. Our evaluation of applying thread-level speculation to client-server applications resulted in substantial performance increases, on the order of 3 times for ourinitial model, and 21 times for the second.
M3 - Article in proceedings
SN - 1932415610, 1932415602
VL - 5
SP - 920
EP - 927
BT - Proceedings of International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA'05)
PB - CSREA Press
ER -
ID: 164443985