ReCoTOS: A Methodology for Vectorization-based Resource-saving Computing Task Optimization

2020 61st International Scientific Conference on Information Technology and Management Science of Riga Technical University (ITMS 2020): Proceedings 2020
Jānis Kampars, Jānis Irbe, Ģirts Kalniņš, Guntis Mosāns, Rasa Gulbe, Krišjānis Pinka

The continuously growing amount of data and complexity of corresponding computing tasks require an adequate computing capacity and software. Processor manufacturers no longer try to gain a significant processor clock speed increase and performance improvements are being provided by other means, such as vectorization which allows to execute Single Instruction Multiple Data (SIMD) operations. Unfortunately, this has introduced new challenges for software engineering industry since writing software that is able to make full use of the capabilities of a modern processor is not a trivial task. The article presents a methodology, namely ReCoTOS, for vectorization-based Resource-saving computing task optimization and experimentally demonstrates its applicability by optimizing a software correlator for space debris data processing.

Atslēgas vārdi
software optimization, vectorization, SIMD
DOI
10.1109/ITMS51158.2020.9259289
Hipersaite
https://ieeexplore.ieee.org/document/9259289

Kampars, J., Irbe, J., Kalniņš, Ģ., Mosāns, G., Gulbe, R., Pinka, K. ReCoTOS: A Methodology for Vectorization-based Resource-saving Computing Task Optimization. No: 2020 61st International Scientific Conference on Information Technology and Management Science of Riga Technical University (ITMS 2020): Proceedings, Latvija, Riga, 15.-16. oktobris, 2020. Piscataway: IEEE, 2020, 1.-6.lpp. ISBN 978-1-7281-9106-5. e-ISBN 978-1-7281-9105-8. Pieejams: doi:10.1109/ITMS51158.2020.9259289

Publikācijas valoda
English (en)