ATF: A Generic Auto-Tuning Framework. We describe the Auto-Tuning Framework (ATF) - a novel generic approach for automatic program optimization by choosing the most suitable values of program parameters, such as number of parallel threads, tile sizes, etc. Our framework combines four advantages over the state-of-the-art autotuning: i) it is generic regarding the programming language, application domain, tuning objective (e.g., high performance and/or low energy consumption), and search technique; ii) it can auto-tune a broader class of applications by allowing tuning parameters to be interdependent, e.g., when one parameter is divisible by another parameter; iii) it allows tuning parameters with substantially larger ranges by implementing an optimized search space generation process; and iv) its interface is arguably simpler than the interfaces of current auto-tuning frameworks. We demonstrate ATF’s efficacy by comparing it to the state-of-the-art auto-tuning approaches OpenTuner and CLTune, showing better tuning results with less programmer’s effort.
References in zbMATH (referenced in 1 article )
Showing result 1 of 1.
- Gadioli, Davide; Vitali, Emanuele; Palermo, Gianluca; Silvano, Cristina: mARGOt: a dynamic autotuning framework for self-aware approximate computing (2019)