2026-06-03T00:42:57Zhttps://riuma.uma.es/rest/oai/request

oai:riuma.uma.es:10630/263872026-02-03T11:04:38Zcom_10630_2254col_10630_37953

Quislant-del-Barrio, Ricardo Fernández-Vega, Iván Gutiérrez-Carrasco, Eladio Damián Plata-González, Óscar Guillermo 2023-04-24T10:58:00Z 2023-04-24T10:58:00Z 2023 Quislant, R., Fernandez, I., Gutierrez, E. et al. Time series analysis acceleration with advanced vectorization extensions. J Supercomput (2023). https://doi.org/10.1007/s11227-023-05060-2 https://hdl.handle.net/10630/26387 10.1007/s11227-023-05060-2 Time series analysis is an important research topic and a key step in monitoring and predicting events in many felds. Recently, the Matrix Profle method, and particularly two of its Euclidean-distance-based implementations—SCRIMP and SCAMP—have become the state-of-the-art approaches in this feld. Those algorithms bring the possibility of obtaining exact motifs and discords from a time series, which can be used to infer events, predict outcomes, detect anomalies and more. While matrix profle is embarrassingly parallelizable, we fnd that auto-vectorization techniques fail to fully exploit the SIMD capabilities of modern CPU architectures. In this paper, we develop custom-vectorized SCRIMP and SCAMP implementations based on AVX2 and AVX-512 extensions, which we combine with multithreading techniques aimed at exploiting the potential of the underneath architectures. Our experimental evaluation, conducted using real data, shows a performance improvement of more than 4× with respect to the auto-vectorization. eng http://creativecommons.org/licenses/by/4.0/ open access Atribución 4.0 Internacional Proceso de vectores (Informática) Time series analysis acceleration with advanced vectorization extensions journal article