An adaptive task creation strategy for work-stealing scheduling

Lei Wang; Huimin Cui; Yuelu Duan; Fang Lu; Xiaobing Feng; Pen Chung Yew

doi:10.1145/1772954.1772992

An adaptive task creation strategy for work-stealing scheduling

Lei Wang, Huimin Cui, Yuelu Duan, Fang Lu, Xiaobing Feng, Pen Chung Yew

Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

28 Scopus citations

Abstract

Work-stealing is a key technique in many multi-threading programming languages to get good load balancing. The current work-stealing techniques have a high implementation overhead in some applications and require a large amount of memory space for data copying to assure correctness. They also cannot handle many application programs that have an unbalanced call tree or have no definitive working sets. In this paper, we propose a new adaptive task creation strategy, called AdaptiveTC, which supports effective work-stealing schemes and also handles the above mentioned problems effectively. As shown in some experimental results, AdaptiveTC runs 2.71x faster than Cilk and 1.72x faster than Tascell for the 16-queen problem with 8 threads.

Original language	English (US)
Title of host publication	Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization
Pages	266-277
Number of pages	12
DOIs	https://doi.org/10.1145/1772954.1772992
State	Published - 2010
Event	8th International Symposium on Code Generation and Optimization, CGO 2010 - Toronto, ON, Canada Duration: Apr 24 2010 → Apr 28 2010

Publication series

Name	Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization

Other

Other	8th International Symposium on Code Generation and Optimization, CGO 2010
Country/Territory	Canada
City	Toronto, ON
Period	4/24/10 → 4/28/10

Keywords

adaptive
backtracking search
task granularity
work-stealing

Access

10.1145/1772954.1772992

OpenUrl availability

Full text

Cite this

Wang, L., Cui, H., Duan, Y., Lu, F., Feng, X., & Yew, P. C. (2010). An adaptive task creation strategy for work-stealing scheduling. In Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization (pp. 266-277). (Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization). https://doi.org/10.1145/1772954.1772992

An adaptive task creation strategy for work-stealing scheduling. / Wang, Lei; Cui, Huimin; Duan, Yuelu et al.
Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization. 2010. p. 266-277 (Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Wang, L, Cui, H, Duan, Y, Lu, F, Feng, X & Yew, PC 2010, An adaptive task creation strategy for work-stealing scheduling. in Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization. Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization, pp. 266-277, 8th International Symposium on Code Generation and Optimization, CGO 2010, Toronto, ON, Canada, 4/24/10. https://doi.org/10.1145/1772954.1772992

@inproceedings{167159d3a0cc4ad2b5a5a60f7351b14e,

title = "An adaptive task creation strategy for work-stealing scheduling",

abstract = "Work-stealing is a key technique in many multi-threading programming languages to get good load balancing. The current work-stealing techniques have a high implementation overhead in some applications and require a large amount of memory space for data copying to assure correctness. They also cannot handle many application programs that have an unbalanced call tree or have no definitive working sets. In this paper, we propose a new adaptive task creation strategy, called AdaptiveTC, which supports effective work-stealing schemes and also handles the above mentioned problems effectively. As shown in some experimental results, AdaptiveTC runs 2.71x faster than Cilk and 1.72x faster than Tascell for the 16-queen problem with 8 threads.",

keywords = "adaptive, backtracking search, task granularity, work-stealing",

author = "Lei Wang and Huimin Cui and Yuelu Duan and Fang Lu and Xiaobing Feng and Yew, {Pen Chung}",

year = "2010",

doi = "10.1145/1772954.1772992",

language = "English (US)",

isbn = "9781605586359",

series = "Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization",

pages = "266--277",

booktitle = "Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization",

}

TY - GEN

T1 - An adaptive task creation strategy for work-stealing scheduling

AU - Wang, Lei

AU - Cui, Huimin

AU - Duan, Yuelu

AU - Lu, Fang

AU - Feng, Xiaobing

AU - Yew, Pen Chung

PY - 2010

Y1 - 2010

N2 - Work-stealing is a key technique in many multi-threading programming languages to get good load balancing. The current work-stealing techniques have a high implementation overhead in some applications and require a large amount of memory space for data copying to assure correctness. They also cannot handle many application programs that have an unbalanced call tree or have no definitive working sets. In this paper, we propose a new adaptive task creation strategy, called AdaptiveTC, which supports effective work-stealing schemes and also handles the above mentioned problems effectively. As shown in some experimental results, AdaptiveTC runs 2.71x faster than Cilk and 1.72x faster than Tascell for the 16-queen problem with 8 threads.

AB - Work-stealing is a key technique in many multi-threading programming languages to get good load balancing. The current work-stealing techniques have a high implementation overhead in some applications and require a large amount of memory space for data copying to assure correctness. They also cannot handle many application programs that have an unbalanced call tree or have no definitive working sets. In this paper, we propose a new adaptive task creation strategy, called AdaptiveTC, which supports effective work-stealing schemes and also handles the above mentioned problems effectively. As shown in some experimental results, AdaptiveTC runs 2.71x faster than Cilk and 1.72x faster than Tascell for the 16-queen problem with 8 threads.

KW - adaptive

KW - backtracking search

KW - task granularity

KW - work-stealing

UR - http://www.scopus.com/inward/record.url?scp=77953990150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77953990150&partnerID=8YFLogxK

U2 - 10.1145/1772954.1772992

DO - 10.1145/1772954.1772992

M3 - Conference contribution

AN - SCOPUS:77953990150

SN - 9781605586359

T3 - Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization

SP - 266

EP - 277

BT - Proceedings of the 2010 CGO - The 8th International Symposium on Code Generation and Optimization

T2 - 8th International Symposium on Code Generation and Optimization, CGO 2010

Y2 - 24 April 2010 through 28 April 2010

ER -

An adaptive task creation strategy for work-stealing scheduling

Abstract

Publication series

Other

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this