Investigating the Scalability of Parallel DTW Distance Computations of Stock Price Time Series
Resumo
Time-series processing and analysis play a fundamental role in several scientific and industrial domains, particularly in identifying patterns and relationships in temporal data. In the financial domain, similarity analysis between assets can reveal correlated behavior and support decision-making. However, computing Dynamic Time Warping (DTW) distances is computationally expensive when applied to large collections of time series. This work investigates the scalability of parallel implementations based on OpenMP, MPI, and the hybrid MPI+OpenMP model to accelerate DTW distance computation on high-performance computing systems. Different communication strategies, workload distribution schemes, and task granularities are evaluated through experiments conducted on a multiprocessor and a multicore cluster, considering speedup and parallel efficiency as performance metrics. Results demonstrate that in general parallel computing substantially increases the practicality of DTW-based similarity analysis for large financial datasets, with the hybrid MPI+OpenMP implementation obtaining over 90% efficiency. Furthermore, the optimization strategies and scalability insights presented in this work can be extended to other domains that rely on large-scale time-series processing, such as climate modeling, bioinformatics, and sensor networks.
Referências
Chan, G. Y.-Y., Nonato, L. G., Palpanas, T., Silva, C. T., and Freire, J. (2026). Tivy: Time series visual summary for scalable visualization. IEEE Transactions on Visualization and Computer Graphics, 32(1):232–242.
de Jesus, L. C., Fernández-Navarro, F., and Carbonero-Ruz, M. (2025). Enhancing financial time series forecasting through topological data analysis. Neural Comput Applications, 37:6527–6545.
Hamerski, J. C., Abich, G., Reis, R., Ost, L., and Amory, A. (2017). Publish-subscribe programming for a noc-based multiprocessor system-on-chip. In 2017 IEEE International Symposium on Circuits and Systems (ISCAS), pages 1–4.
Huang, S., Dai, G., Sun, Y., Wang, Z., Wang, Y., and Yang, H. (2013). Dtw-based subsequence similarity search on amd heterogeneous computing platform. In 2013 IEEE 10th International Conference on High Performance Computing and Communications 2013 IEEE International Conference on Embedded and Ubiquitous Computing, pages 1054–1063.
Meert, W., Hendrickx, K., Van Craenendonck, T., Robberechts, P., Blockeel, H., and Davis, J. (2020). Dtaidistance.
Miniakhmetov, R., Movchan, A., and Zymbler, M. (2015). Accelerating time series subsequence matching on the intel xeon phi many-core coprocessor. In 2015 38th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), pages 1399–1404.
Ninama, K., Patel, J., K K, G., Reddy, M. R. V. S. R. S., and Bhowmik, B. (2025). Performance analysis of hybrid mpi and openmp on smith-waterman algorithm. In 2025 3rd International Conference on Intelligent Systems, Advanced Computing and Communication (ISACC), pages 887–892.
Radovanović, A., Ye, X., Milanović, J. V., Milosavljević, N., and Storchi, R. (2020). Application of the k-medoids partitioning algorithm for clustering of time series data. In 2020 IEEE PES Innovative Smart Grid Technologies Europe (ISGT-Europe), pages 645–649.
Saklani, R., Purohit, K., Vats, S., Sharma, V., Kukreja, V., and Yadav, S. P. (2023). Multicore implementation of k-means clustering algorithm. In 2023 2nd International Conference on Applied Artificial Intelligence and Computing (ICAAIC), pages 171–175.
Sakoe, H. and Chiba, S. (1978). Dynamic programming algorithm optimization for spoken word recognition. IEEE Transactions on Acoustics, Speech, and Signal Processing, 26(1):43–49.
Sigaki, H. Y. D., Perc, M., and Ribeiro, H. V. (2019). Clustering patterns in efficiency and the coming-of-age of the cryptocurrency market. Scientific Reports, 9(1):1440.
Vintsyuk, T. K. (1968). Speech discrimination by dynamic programming. Cybernetics, 4(1):52–57.
Winarti, T., Widodo, E., Handayni, S., and Nugroho, A. (2024). Utilizing pearson correlation matrix to identify negative correlations among cryptocurrencies. In 2024 3rd International Conference on Creative Communication and Innovative Technology (ICCIT), pages 1–7.
Yang, D., Shaw, T., and Tsai, T. (2022). A study of parallelizable alternatives to dynamic time warping for aligning long sequences. IEEE/ACM Trans. Audio, Speech and Lang. Proc., 30:2117–2127.
Zhang, R., Li, H., and Wang, J. (2021). Index tracking based on dynamic time warping and constrained k-medoids clustering. In 2021 11th International Conference on Intelligent Control and Information Processing (ICICIP), pages 352–359.
Zhu, B., Jiang, Y., Gu, M., and Deng, Y. (2021). A gpu acceleration framework for motif and discord based pattern mining. IEEE Transactions on Parallel and Distributed Systems, 32(8):1987–2004.
