Article

Scheduling dynamic OpenMP applications over multicore architectures

Authors:

François Broquedis,

François Diakhaté,

Samuel Thibault,

Olivier Aumage,

Raymond Namyst,

Pierre-André WacrenierAuthors Info & Claims

IWOMP'08: Proceedings of the 4th international conference on OpenMP in a new era of parallelism

Pages 170 - 180

Published: 12 May 2008 Publication History

Abstract

Approaching the theoretical performance of hierarchical multicoremachines requires a very careful distribution of threads and dataamong the underlying non-uniform architecture in order to minimizecache misses and NUMA penalties. While it is acknowledged thatOpenMP can enhance the quality of thread scheduling on such architecturesin a portable way, by transmitting precious information aboutthe affinities between threads and data to the underlying runtime system,most OpenMP runtime systems are actually unable to efficiently supporthighly irregular, massively parallel applications on NUMA machines.

In this paper, we present a thread scheduling policy suited to theexecution of OpenMP programs featuring irregular and massive nestedparallelism over hierarchical architectures. Our policy enforces a distributionof threads that maximizes the proximity of threads belonging tothe same parallel region, and uses a NUMA-aware work stealing strategywhen load balancing is needed. It has been developed as a plug-in tothe forestGOMP OpenMP platform [TBG+07]. We demonstrate theefficiency of our approach with a highly irregular recursive OpenMP programresulting from the generic parallelization of a surface reconstructionapplication. We achieve a speedup of 14 on a 16-core machine with noapplication-level optimization.

References

[1]

Ayguade, E., Copty, N., Duranl, A., Hoeflinger, J., Lin, Y., Massaioli, F.,Su, E., Unnikrishnan, P., Zhang, G.: A proposal for task parallelism inOpenMP. In: Third International Workshop on OpenMP (IWOMP 2007),Beijing, China (2007).

Digital Library

[2]

Ayguade, E., Gonzalez, M., Martorell, X., Jost, G.: Employing NestedOpenMP for the Parallelization of Multi-Zone Computational Fluid DynamicsApplications. In: 18th International Parallel and Distributed ProcessingSymposium (IPDPS) (2004).

[3]

an Mey, D., Sarholz, S., Terboven, C.: Nested Parallelization withOpenMP. Parallel Computing 35(5), 459-476 (2007).

Digital Library

[4]

Balart, J., Duran, A., Gonzàlez, M., Martorell, X., Ayguadé, E., Labarta,J.: Nanos mercurium: A research compiler for openmp. In: European Workshopon OpenMP (EWOMP) (October 2004).

[5]

Blikberg, R., Sørevik, T.: Load balancing and OpenMP implementation ofnested parallelism. Parallel Computing, 31(10-12):984-998 (October 2005).

Digital Library

[6]

Chapman, B.M., Huang, L., Jin, H., Jost, G., de Supinski, B.R.: Extendingopenmp worksharing directives for multithreading. In: EuroPar 2006Parallel Processing (2006).

[7]

Duran, A., Gonzàles, M., Corbalán, J.: Automatic Thread Distribution forNested Parallelism in OpenMP. In: 19th ACM International Conference onSupercomputing, Cambridge, MA, USA, June 2005, pp. 121-130 (2005).

Digital Library

[8]

Duran, A., Silvera, R., Corbalán, J., Labarta, J.: Runtime adjustment ofparallel nested loops. In: Chapman, B.M. (ed.) WOMPAT 2004. LNCS,vol. 3349, Springer, Heidelberg (2005).

Digital Library

[9]

Frigo, M., Leiserson, C.E., Randall, K.H.: The Implementation of the Cilk-5 Multithreaded Language. In: ACM SIGPLAN Conference on ProgrammingLanguage Design and Implementation (PLDI), Montreal, Canada (June 1998).

Digital Library

[10]

GOMP - An OpenMP implementation for GCC,http://gcc.gnu.org/projects/gomp/

[11]

Gonzalez, M., Oliver, J., Martorell, X., Ayguade, E., Labarta, J., Navarro,N.: OpenMP Extensions for Thread Groups and Their Run-Time Support.In: Languages and Compilers for Parallel Computing, Springer, Heidelberg(2001).

Digital Library

[12]

Gao, G.R., Sterling, T., Stevens, R., Hereld, M., Zhu, W.: Hierarchicalmultithreading: programming model and system software. In: 20th InternationalParallel and Distributed Processing Symposium (IPDPS) (April 2006).

Digital Library

[13]

Gerndt, A., Sarholz, S., Wolter, M., an Mey, D., Bischof, C., Kuhlen, T.:Nested OpenMP for Efficient Computation of 3D Critical Points in Multi-Block CFD Datasets. In: Super Computing (November 2006).

Digital Library

[14]

Hadjidoukas, P.E., Dimakopoulos, V.V.: Nested Parallelism in the OMPiOpenMP/C compiler. In: EuroPar, Rennes,France, July 2007, ACM, NewYork (2007).

Digital Library

[15]

Karlsson, S.: An Introduction to Balder - An OpenMP Run-time Libraryfor Clusters of SMPs. In: International Workshop on OpenMP (IWOMP)(June 2005).

Digital Library

[16]

Martorell, X., Ayguadé, E., Navarro, N., Corbalán, J., González, M.,Labarta, J.: Thread Fork/Join Techniques for Multi-Level Parallelism Exploitationin NUMA Multiprocessors. In: International Conference on SuperComputing,pp. 294-301. ACM Press, New York (1999).

Digital Library

[17]

Ohtake, Y., Belyaev, A., Alexa, M., Turk, G., Seidel, H.-P.: Multi-level partition of unity implicits. ACM Trans. Graph. 22(3), 463-470 (2003).

Digital Library

[18]

Su, E., Tian, X., Haab, M.G.G., Shah, S., Petersen, P.: Compiler Supportof the Workqueuing Execution Model for Intel SMP Architectures. In:European Workshop on OpenMP (EWOMP) (October 2004).

[19]

Thibault, S., Broquedis, F., Goglin, B., Namyst, R., Wacrenier, P.-A.: AnEfficient OpenMP Runtime System for Hierarchical Architectures. In: InternationalWorkshop on OpenMP (IWOMP), Beijing,China, June 2007,pp. 148-159 (2007).

Digital Library

[20]

Tian, X., Girkar, M., Bik, A., Saito, H.: Practical Compiler Techniques onEfficient Multithreaded Code Generation for OpenMP Programs. Comput.J. 48(5), 588-601 (2005).

Digital Library

[21]

Tian, X., Girkar, M., Shah, S., Armstrong, D., Su, E., Petersen, P.: Compilerand Runtime Support for Running OpenMP Programs on PentiumandItanium-Architectures. In: Eighth International Workshop on High-Level Parallel Programming Models and Supportive Environments, April2003, pp. 47-55 (2003).

Digital Library

[22]

Tian, X., Hoeflinger, J.P., Haab, G., Chen, Y.-K., Girkar, M., Shah, S.: Acompiler for exploiting nested parallelism in OpenMP programs. ParallelComput. 31(10-12), 960-983 (2005).

Digital Library

[23]

Tanaka, Y., Taura, K., Sato, M., Yonezawa, A.: Performance evaluationof openmp applications with nested parallelism. In: Languages, Compilers,and Run-Time Systems for Scalable Computers, pp. 100-112 (2000).

Digital Library

Cited By

Abadi MKeidar-Barner SPidan DVeksler T(2019)Verifying Parallel Code After Refactoring Using Equivalence CheckingInternational Journal of Parallel Programming10.1007/s10766-017-0548-447:1(59-73)Online publication date: 1-Feb-2019
https://dl.acm.org/doi/10.1007/s10766-017-0548-4
Marzolla MD'Angelo GGarro AD'Ambrogio A(2017)Parallel sort-based matching for data distribution management on shared-memory multiprocessorsProceedings of the 21st International Symposium on Distributed Simulation and Real Time Applications10.5555/3199858.3199860(1-8)Online publication date: 18-Oct-2017
https://dl.acm.org/doi/10.5555/3199858.3199860
León EJacob B(2017)mpibindProceedings of the International Symposium on Memory Systems10.1145/3132402.3132415(262-264)Online publication date: 2-Oct-2017
https://dl.acm.org/doi/10.1145/3132402.3132415
Show More Cited By

Recommendations

An Efficient OpenMP Runtime System for Hierarchical Architectures
IWOMP '07: Proceedings of the 3rd international workshop on OpenMP: A Practical Programming Model for the Multi-Core Era

Exploiting the full computational power of always deeper hierarchical multiprocessor machines requires a very careful distribution of threads and data among the underlying non-uniform architecture. The emergence of multi-core chips and NUMA machines ...
A compiler for exploiting nested parallelism in OpenMP programs
OpenMp

This paper presents the design and implementation of a parallelization framework and OpenMP runtime support in Intel C++ & Fortran compilers for exploiting nested parallelism in applications using OpenMP pragmas or directives. We conduct the performance ...
Performance Gaps between OpenMP and OpenCL for Multi-core CPUs
ICPPW '12: Proceedings of the 2012 41st International Conference on Parallel Processing Workshops

OpenCL and OpenMP are the most commonly used programming models for multi-core processors. They are also fundamentally different in their approach to parallelization. In this paper, we focus on comparing the performance of OpenCL and OpenMP. We select ...

Comments

Information & Contributors

Information

Published In

cover image Guide Proceedings

IWOMP'08: Proceedings of the 4th international conference on OpenMP in a new era of parallelism

May 2008

191 pages

ISBN:354079560X

Editors:
Rudolf Eigenmann
Purdue University, School of Electrical and Computer Engineering, West Lafayette, IN
,
Bronis R. De Supinski
Computation Directorate, Lawrence Livermore National Laboratory, Livermore, CA

Publisher

Springer-Verlag

Berlin, Heidelberg

Publication History

Published: 12 May 2008

Author Tags

Qualifiers

Article

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

7
Total Citations
View Citations
11
Total Downloads

Downloads (Last 12 months)0
Downloads (Last 6 weeks)0

Reflects downloads up to 05 Feb 2025

Other Metrics

View Author Metrics

Citations

Cited By

Abadi MKeidar-Barner SPidan DVeksler T(2019)Verifying Parallel Code After Refactoring Using Equivalence CheckingInternational Journal of Parallel Programming10.1007/s10766-017-0548-447:1(59-73)Online publication date: 1-Feb-2019
https://dl.acm.org/doi/10.1007/s10766-017-0548-4
Marzolla MD'Angelo GGarro AD'Ambrogio A(2017)Parallel sort-based matching for data distribution management on shared-memory multiprocessorsProceedings of the 21st International Symposium on Distributed Simulation and Real Time Applications10.5555/3199858.3199860(1-8)Online publication date: 18-Oct-2017
https://dl.acm.org/doi/10.5555/3199858.3199860
León EJacob B(2017)mpibindProceedings of the International Symposium on Memory Systems10.1145/3132402.3132415(262-264)Online publication date: 2-Oct-2017
https://dl.acm.org/doi/10.1145/3132402.3132415
Busse ASchönherr JDiener MNavaux PHeiß HWainwright RCorchado JBechini AHong J(2015)Partial coscheduling of virtual machines based on memory access patternsProceedings of the 30th Annual ACM Symposium on Applied Computing10.1145/2695664.2695736(2033-2038)Online publication date: 13-Apr-2015
https://dl.acm.org/doi/10.1145/2695664.2695736
Di Biagio ASpeziale EAgosta G(2011)Exploiting thread-data affinity in OpenMP with data access patternsProceedings of the 17th international conference on Parallel processing - Volume Part I10.5555/2033345.2033370(230-241)Online publication date: 29-Aug-2011
https://dl.acm.org/doi/10.5555/2033345.2033370
Yan JHe JHan WChen WZheng W(2010)How OpenMP applications get more benefit from many-core eraProceedings of the 6th international conference on Beyond Loop Level Parallelism in OpenMP: accelerators, Tasking and more10.1007/978-3-642-13217-9_7(83-95)Online publication date: 14-Jun-2010
https://dl.acm.org/doi/10.1007/978-3-642-13217-9_7
Broquedis FFurmento NGoglin BNamyst RWacrenier P(2009)Dynamic Task and Data Placement over NUMA ArchitecturesProceedings of the 5th International Workshop on OpenMP: Evolving OpenMP in an Age of Extreme Parallelism10.1007/978-3-642-02303-3_7(79-92)Online publication date: 22-May-2009
https://dl.acm.org/doi/10.1007/978-3-642-02303-3_7

View Options

View options

Figures

Tables

Media

View Table of Conten