Towards Unifying OpenMP Under the Task-Parallel Paradigm

Implementation and Performance of the taskloop Construct
  • Artur PodobasEmail author
  • Sven Karlsson
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9903)


OpenMP 4.5 introduced a task-parallel version of the classical thread-parallel for-loop construct: the taskloop construct. With this new construct, programmers are given the opportunity to choose between the two parallel paradigms to parallelize their for loops. However, it is unclear where and when the two approaches should be used when writing efficient parallel applications.

In this paper, we explore the taskloop construct. We study performance differences between traditional thread-parallel for loops and the new taskloop directive. We introduce an efficient implementation and compare our implementation to other taskloop implementations using micro- and kernel-benchmarks, as well as an application. We show that our taskloop implementation on average results in a 3.2 % increase in peak performance when compared against corresponding parallel-for loops.


Execution Time Load Balance Critical Path Single Iteration Iteration Space 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.



We acknowledge the reviewers for their suggestions in making this paper better. The research leading to these results has received funding from the ARTEMIS Joint Undertaking under grant agreement number 332913 for project COPCAMS.


  1. 1.
    Acar, U.A., Blelloch, G.E., Blumofe, R.D.: The data locality of work stealing. In: Proceedings of the Annual ACM Symposium on Parallel Algorithms and Architectures, pp. 1–12. ACM (2000)Google Scholar
  2. 2.
    Aslot, V., Domeika, M., Eigenmann, R., Gaertner, G., Jones, W.B., Parady, B.: SPEComp: a new benchmark suite for measuring parallel computer performance. In: Eigenmann, R., Voss, M.J. (eds.) WOMPAT 2001. LNCS, vol. 2104, pp. 1–10. Springer, Heidelberg (2001)CrossRefGoogle Scholar
  3. 3.
    Ayguadé, E., Copty, N., Duran, A., Hoeflinger, J., Lin, Y., Massaioli, F., Teruel, X., Unnikrishnan, P., Zhang, G.: The design of OpenMP tasks. IEEE Trans. Parallel Distrib. Syst. 20(3), 404–418 (2009)CrossRefGoogle Scholar
  4. 4.
    Bienia, C., Li, K.: PARSEC 2.0: a new benchmark suite for chip-multiprocessors. In: Proceedings of the Annual Workshop on Modeling, Benchmarking and Simulation, vol. 2011 (2009)Google Scholar
  5. 5.
    Bohme, D., Wolf, F., Supinski, D., Bronis, R., Schulz, M., Geimer, M.: Scalable critical-path based performance analysis. In: Proceedings of Parallel & Distributed Processing Symposium, pp. 1330–1340. IEEE (2012)Google Scholar
  6. 6.
    Bonnichsen, L., Podobas, A.: Using transactional memory to avoid blocking in OpenMP synchronization directives. In: Terboven, C., et al. (eds.) IWOMP 2015. LNCS, vol. 9342, pp. 149–161. Springer, Heidelberg (2015). doi: 10.1007/978-3-319-24595-9_11 CrossRefGoogle Scholar
  7. 7.
    Che, S., Boyer, M., Meng, J., Tarjan, D., Sheaffer, J.W., Lee, S.-H., Skadron, K.: Rodinia: a benchmark suite for heterogeneous computing. In: Proceedings of IEEE International Symposium on Workload Characterization, pp. 44–54. IEEE (2009)Google Scholar
  8. 8.
    Duran, A., Ayguadé, E., Badia, R.M., Labarta, J., Martinell, L., Martorell, X., Planas, J.: OmpSs: a proposal for programming heterogeneous multi-core architectures. Parallel Process. Lett. 21(02), 173–193 (2011)MathSciNetCrossRefGoogle Scholar
  9. 9.
    Goldstein, S.C., Schauser, K.E., Culler, D.E.: Lazy threads: implementing a fast parallel call. J. Parallel Distrib. Comput. 37(1), 5–20 (1996)CrossRefGoogle Scholar
  10. 10.
    González, C.H., Fraguela, B.B.: A generic algorithm template for divide-and-conquer in multicore systems. In: Proceedings of IEEE International Conference on High Performance Computing and Communications, pp. 79–88. IEEE (2010)Google Scholar
  11. 11.
    Kumar, P.: Cache oblivious algorithms. In: Petreschi, R., Persiano, G., Silvestri, R. (eds.) CIAC 2003. LNCS, vol. 2653, pp. 193–212. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  12. 12.
    Leiserson, C.E.: The Cilk++ concurrency platform. J. Supercomput. 51(3), 244–257 (2010)CrossRefGoogle Scholar
  13. 13.
    Mohr, E., Kranz, D.A., Halstead Jr., R.H.: Lazy task creation: a technique for increasing the granularity of parallel programs. IEEE Trans. Parallel Distrib. Syst. 2(3), 264–280 (1991)CrossRefGoogle Scholar
  14. 14.
    Podobas, A., Brorsson, M., Vlassov, V.: TurboBŁYSK: scheduling for improved data-driven task performance with fast dependency resolution. In: DeRose, L., Supinski, B.R., Olivier, S.L., Chapman, B.M., Müller, M.S. (eds.) IWOMP 2014. LNCS, vol. 8766, pp. 45–57. Springer, Heidelberg (2014)Google Scholar
  15. 15.
    Polychronopoulos, C.D., Kuck, D.J.: Guided self-scheduling: a practical scheduling scheme for parallel supercomputers. IEEE Trans. Comput. 100(12), 1425–1439 (1987)CrossRefGoogle Scholar
  16. 16.
    Tzen, H.T., Ni, L.M.: Trapezoid self-scheduling: a practical scheduling scheme for parallel compilers. IEEE Trans. Parallel Distrib. Syst. 4(1), 87–98 (1993)CrossRefGoogle Scholar
  17. 17.
    Zhang, Y., Burcea, M., Cheng, V., Ho, R., Voss, M.: An adaptive OpenMP loop scheduler for hyperthreaded SMPs. In: Proceedings of International Conference on Parallel and Distributed Computing (and Communications) Systems, pp. 256–263 (2004)Google Scholar
  18. 18.
    Zhang, Y., Voss, M., Rogers, E.S.: Runtime empirical selection of loop schedulers on hyperthreaded smps. In: Proceedings of International Parallel and Distributed Processing Symposium, p. 44b. IEEE (2005)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.Technical University of DenmarkKongens LyngbyDenmark

Personalised recommendations