GPU Extended Stock Market Software Architecture

  • Alisa KrstovaEmail author
  • Marjan Gusev
  • Vladimir Zdraveski
Conference paper
Part of the Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering book series (LNICST, volume 283)


We propose a stock market software architecture extended by a graphics processing unit, which employs parallel programming paradigm techniques to optimize long-running tasks like computing daily trends and performing statistical analysis of stock market data in real-time. The system uses the ability of Nvidia’s CUDA parallel computation application programming interface (API) to integrate with traditional web development frameworks. The web application offers extensive statistics and stocks’ information which is periodically recomputed through scheduled batch jobs or calculated in real-time. To illustrate the advantages of using many-core programming, we explore several use-cases and evaluate the improvement in performance and speedup obtained in comparison to the traditional approach of executing long-running jobs on a central processing unit (CPU).


Stock market GPU Parallel programming CUDA 


  1. 1.
    Peachavanish, R.: Stock selection and trading based on cluster analysis of trend and momentum indicators. In: Proceedings of the International MultiConference of Engineers and Computer Scientists, vol. 1, pp. 317–321 (2016)Google Scholar
  2. 2.
    Marjanovic, B.: Huge stock market dataset. Accessed 02 May 2018
  3. 3.
    Golan, R.H., Ziarko, W.: A methodology for stock market analysis utilizing rough set theory. In: Computational Intelligence for Financial Engineering: Proceedings of the IEEE/IAFE 1995, pp. 32–40. IEEE (1995)Google Scholar
  4. 4.
    Marketwatch - an online virtual stock market simulator. Accessed 02 May 2018
  5. 5.
    Gariney, V.: Statistical analysis for daily forecast of stock prices (2002)Google Scholar
  6. 6.
    Chang, D.-J., Desoky, A.H., Ouyang, M., Rouchka, E.C.: Compute pairwise Manhattan distance and Pearson correlation coefficient of data points with GPU. In: 2009 10th ACIS International Conference on Software Engineering, Artificial Intelligences, Networking and Parallel/Distributed Computing, pp. 501–506 (2009)Google Scholar
  7. 7.
    Kijsipongse, E., Suriya, U., Ngamphiw, C., Tongsima, S.: Efficient large Pearson correlation matrix computing using hybrid MPI/CUDA. In: 2011 Eighth International Joint Conference on Computer Science and Software Engineering (JCSSE), pp. 237–241, May 2011Google Scholar
  8. 8.
    Klöckner, A., Pinto, N., Catanzaro, B., Lee, Y., Ivanov, P., Fasih, A.: GPU scripting and code generation with PyCUDA. In: GPU Computing Gems Jade Edition, pp. 373–385. Elsevier (2011)Google Scholar
  9. 9.
    Mu, Q., Cui, L., Song, Y.: The implementation and optimization of Bitonic sort algorithm based on CUDA, CoRR, vol. abs/1506.01446 (2015).
  10. 10.
    Ionescu, M.F., Schauser, K.E.: Optimizing parallel Bitonic sort. In: Parallel Processing Symposium: Proceedings, 11th International, pp. 303–309. IEEE (1997)Google Scholar
  11. 11.
    NVIDIA Corporation: Compute capabilities. Accessed 12 May 2018
  12. 12.
    NVIDIA Corporation, Gameworks Documentation, “Achieved occupancy”. Accessed 12 May 2018
  13. 13.
    Connors, T.A., Qasem, A.: Automatically selecting profitable thread block sizes for accelerated kernels. In: 2017 IEEE 19th International Conference on High Performance Computing and Communications; IEEE 15th International Conference on Smart City; IEEE 3rd International Conference on Data Science and Systems (HPCC/SmartCity/DSS), pp. 442–449, December 2017Google Scholar

Copyright information

© ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2019

Authors and Affiliations

  • Alisa Krstova
    • 1
    Email author
  • Marjan Gusev
    • 1
  • Vladimir Zdraveski
    • 1
  1. 1.Faculty of Computer Science and EngineeringUniversity Ss. Cyril and MethodiusSkopjeMacedonia

Personalised recommendations