Kernel Tuning Toolkit

Warning

This publication doesn't include Faculty of Economics and Administration. It includes Institute of Computer Science. Official publication website can be found on muni.cz.

Authors	PETROVIČ Filip FILIPOVIČ Jiří
Year of publication	2023
Type	Article in Periodical
Magazine / Source	SoftwareX
MU Faculty or unit	Institute of Computer Science
Citation
Web	https://www.sciencedirect.com/science/article/pii/S235271102300081X
Doi	http://dx.doi.org/10.1016/j.softx.2023.101385
Keywords	Autotuning; GPU optimization; CUDA; OpenCL; Vulkan
Description	Kernel Tuning Toolkit (KTT) is an autotuning framework for CUDA, OpenCL and Vulkan kernels. KTT provides advanced autotuning features such as support for both dynamic (online) and offline tuning, and an ability to tune multiple kernels together with shared tuning parameters. Furthermore, it offers customization features that make integration into larger software suites possible. The framework handles all major steps required for autotuning implementation, including configuration space creation and exploration, kernel code execution and output validation. The public API is available natively in C++ and via bindings in Python.
Related projects:	e-Infrastruktura CZ