Publication: Towards a hierarchical approach for autotuning task-based libraries
Authors
Cámara, Jesús ; Cuenca Muñoz, Antonio Javier ; Boratto, Murilo ; Vicente Jaén, Arturo ; Galindo Garre, Víctor
item.page.secondaryauthor
Facultades de la UMU::Facultad de Informática
item.page.director
Publisher
Springer
publication.page.editor
publication.page.department
DOI
https://doi.org/10.1007/s11227-026-08412-w
item.page.type
info:eu-repo/semantics/article
Description
Abstract
This work proposes a hierarchical approach to reduce the training time of task-based routines by reusing previously obtained autotuning information. This approach has been integrated into a working prototype of Chameleon, a dense linear algebra software whose tile-based routines are executed on the available computational resources by means of a runtime system. The results show that this approach provides a high degree of scalability to the entire self-optimization process, achieving a reduction in training time of up to 80% and an appropriate selection of values for the adjustable parameters.
publication.page.subject
Citation
Cámara, J., Cuenca, J., Galindo, V. et al. An autotuning approach to select the inter-GPU communication library on heterogeneous systems. J Supercomput 81, 283 (2025). https://doi.org/10.1007/s11227-024-06794-3
item.page.embargo
Collections
Ir a Estadísticas
Este ítem está sujeto a una licencia Creative Commons. http://creativecommons.org/licenses/by/4.0/