Advertisement

Accelerating continuous GRASP with a GPU

  • Bruno NogueiraEmail author
  • Eduardo Tavares
  • Jean Araujo
  • Gustavo Callou
Article
  • 22 Downloads

Abstract

This work proposes a GPU-based parallelization for the Continuous GRASP (C-GRASP), a local search metaheuristic for finding cost-efficient solutions to continuous global optimization problems subject to box constraints. C-GRASP has demonstrated competitive performance on several well-known multimodal test functions as well as on difficult real-world problems, hence a GPU parallelization might increase even further the applicability of this metaheuristic. Although GPU parallelizations have been proposed for many metaheuristics, little has been done for C-GRASP. We conduct an extensive set of experiments and compare our proposal with state-of-the-art GPU parallelizations of other metaheuristics, such as Scatter Search and Differential Evolution. We also compare our GPU approach with two other C-GRASP implementations: a sequential version and a multi-core version. Experimental results show our GPU C-GRASP outperforms other GPU-based metaheuristics and the multi-core C-GRASP. Besides, we observed speedups of up to \(154{\times }\) over the sequential version.

Keywords

Continuous optimization Metaheuristics C-GRASP GPU 

Notes

References

  1. 1.
    Resende MG, Ribeiro CC (2016) Optimization by GRASP. Springer, BerlinCrossRefzbMATHGoogle Scholar
  2. 2.
    Hirsch MJ, Meneses C, Pardalos PM, Resende MG (2007) Global optimization by continuous GRASP. Optim Lett 1(2):201–212MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Hirsch MJ, Pardalos PM, Resende MG (2010) Speeding up continuous GRASP. Eur J Oper Res 205(3):507–521CrossRefzbMATHGoogle Scholar
  4. 4.
    Hirsch MJ, Pardalos PM, Resende MG (2006) Sensor registration in a sensor network by continuous GRASP. In: Military Communications Conference, 2006, MILCOM 2006, IEEE. IEEE, pp 1–6Google Scholar
  5. 5.
    Hirsch MJ, Pardalos PM, Resende MG (2009) Solving systems of nonlinear equations with continuous GRASP. Nonlinear Anal Real World Appl 10(4):2000–2006MathSciNetCrossRefzbMATHGoogle Scholar
  6. 6.
    Hirsch MJ, Meneses CN, Pardalos PM, Ragle M, Resende MG (2007) A continuous grasp to determine the relationship between drugs and adverse reactions. In: AIP Conference Proceedings, Vol. 953, AIP, pp 106–121Google Scholar
  7. 7.
    Macharet DG, Neto AA, da Camara Neto VF, Campos MF (2011) Nonholonomic path planning optimization for dubins’ vehicles. In: 2011 IEEE International Conference on Robotics and Automation (ICRA). IEEE, pp 4208–4213Google Scholar
  8. 8.
    Neto JXV, Reynoso-Meza G, Ruppel TH, Mariani VC, dos Santos Coelho L (2017) Solving non-smooth economic dispatch by a new combination of continuous grasp algorithm and differential evolution. Int J Electr Power Energy Syst 84:13–24CrossRefGoogle Scholar
  9. 9.
    Queiroga E, Subramanian A, Lucídio dos Anjos FC (2018) Continuous greedy randomized adaptive search procedure for data clustering. Appl Soft Comput 72:43–55CrossRefGoogle Scholar
  10. 10.
    Nogueira B, Pinheiro RG (2018) A CPU–GPU local search heuristic for the maximum weight clique problem on massive graphs. Comput Oper Res 90:232–248MathSciNetCrossRefzbMATHGoogle Scholar
  11. 11.
    Nogueira B, Pinheiro RG (2019) A GPU based local search algorithm for the unweighted and weighted maximum s-plex problems. Ann Oper Res.  https://doi.org/10.1007/s10479-019-03159-5 Google Scholar
  12. 12.
    Alekseeva E, Mezmaz M, Tuyttens D, Melab N (2017) Parallel multi-core hyper-heuristic GRASP to solve permutation flow-shop problem. Concurr Comput Pract Exp 29(9):e3835CrossRefGoogle Scholar
  13. 13.
    Coelho I, Munhoz P, Ochi L, Souza M, Bentes C, Farias R (2016) An integrated CPU–GPU heuristic inspired on variable neighbourhood search for the single vehicle routing problem with deliveries and selective pickups. Int J Prod Res 54(4):945–962CrossRefGoogle Scholar
  14. 14.
    Melab N, Luong T, Boufaras K, Talbi E.-G (2011) Towards paradisEO-MO-GPU: a framework for GPU-based local search metaheuristics. In: International Work-Conference on Artificial Neural Networks. Springer, pp 401–408Google Scholar
  15. 15.
    Nashed YS, Ugolotti R, Mesejo P, Cagnoni S (2012) libcudaoptimize: an open source library of GPU-based metaheuristics. In: Proceedings of the 14th Annual Conference Companion on Genetic and Evolutionary Computation. ACM, pp 117–124Google Scholar
  16. 16.
    Andrade L, Xavier R, Cabral L, Formiga A, Parallel construction for continuous grasp optimization on GPUs. In: Anais do XLVI Simpósio Brasileiro de Pesquisa Operacional, pp 2393–2404Google Scholar
  17. 17.
    Harris M (2007) Optimizing parallel reduction in CUDA. NVIDIA Dev Technol 2(4):70Google Scholar
  18. 18.
    Jamil M, Yang X-S (2013) A literature survey of benchmark functions for global optimisation problems. Int J Math Model Numer Optim 4(2):150–194zbMATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  • Bruno Nogueira
    • 1
    Email author
  • Eduardo Tavares
    • 2
  • Jean Araujo
    • 3
  • Gustavo Callou
    • 3
  1. 1.Universidade Federal de AlagoasMaceióBrazil
  2. 2.Universidade Federal de PernambucoRecifeBrazil
  3. 3.Universidade Federal Rural de PernambucoRecifeBrazil

Personalised recommendations