Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data – Mgr. et Mgr. Jaroslav Oľha
Mgr. et Mgr. Jaroslav Oľha
Doctoral thesis
Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data
Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data
Abstract:
Moderné vysokovýkonné výpočty často využívajú na dosiahnutie maximálneho výkonu heterogénne hardwarové zdroje -- tento prístup predstavuje zjavné výhody, keďže kombinuje výpočetný výkon viacerých procesorov a umožňuje im väčšiu mieru špecializácie. Takéto výpočty však musia byť naprogramované s ohľadom na daný hardware, čo kladie väčšiu záťaž na programátorov, ktorí musia zabezpečiť, aby ich programy …moreAbstract:
Modern high performance computing applications often rely on heterogeneous hardware resources to achieve maximum performance. This approach presents obvious benefits, combining the processing power of multiple different processors and allowing them to be more specialized. However, since HPC applications typically need to be programmed in a hardware-aware manner to achieve maximum performance, this …more
Language used: English
Date on which the thesis was submitted / produced: 7. 8. 2024
Identifier:
https://is.muni.cz/th/tknpb/
Thesis defence
- Date of defence: 28. 1. 2025
- Supervisor: prof. RNDr. Luděk Matyska, CSc.
- Reader: doc. Ing. Lubomír Říha, PhD., prof. Ana Lucia Varbanescu, PhD
Citation record
ISO 690-compliant citation record:
OĽHA, Jaroslav. \textit{Data-Driven Dynamic Autotuning: Optimizing Autotuning Overhead with Prior Tuning Data}. Online. Doctoral theses, Dissertations. Brno: Masaryk University, Faculty of Informatics. 2024. Available from: https://theses.cz/id/n7cmbs/.
Full text of thesis
Contents of on-line thesis archive
Published in Theses:- světu
Other ways of accessing the text
Institution archiving the thesis and making it accessible: Masarykova univerzita, Fakulta informatikyMasaryk University
Faculty of InformaticsDoctoral programme / field:
Computer Science / Computing Technology and Methodology
Theses on a related topic
-
Modeling the rheology of emulsions: High-performance parallel computing using Graphics Processing Unit (GPU)
Pavel Kupka -
Neuronové sítě v prostředí High Performance Computing
Lukáš Vojáček -
High-Performance Analytics (HPA)
Petr Soukup -
Dynamic autotuning of a fluid dynamics simulation
Petr Babič -
Dynamic autotuning of SpMV kernel in CUSP library
Miroslav Demek -
Dynamic autotuning of SpMV kernel in CUSP library
František Bráblík -
GPU acceleration of the continuous heterogeneity analysis in Xmipp
Dávid Šutor -
GPU acceleration of matrix factorization
Jan Priessnitz