Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures

We have developed several autotuning benchmarks in CUDA that take into account performance-relevant source-code parameters and reach near peak-performance on various GPU architectures. We have used them during the development and evaluation of a search method for tuning space proposed in [1]. With o...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Jana Hozzová, Jiří Filipovič, Amin Nezarat, Jaroslav Ol’ha, Filip Petrovič
Formato:	article
Lenguaje:	EN
Publicado:	Elsevier 2021
Materias:	Auto-tuning Tuning spaces Performance counters Cuda Computer applications to medicine. Medical informatics R858-859.7 Science (General) Q1-390
Acceso en línea:	https://doaj.org/article/18d5ea99c4e041b1adf14cc8b3d5c466
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:18d5ea99c4e041b1adf14cc8b3d5c466
record_format	dspace
spelling	oai:doaj.org-article:18d5ea99c4e041b1adf14cc8b3d5c4662021-11-28T04:33:28ZSearching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures2352-340910.1016/j.dib.2021.107631https://doaj.org/article/18d5ea99c4e041b1adf14cc8b3d5c4662021-12-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S2352340921009069https://doaj.org/toc/2352-3409We have developed several autotuning benchmarks in CUDA that take into account performance-relevant source-code parameters and reach near peak-performance on various GPU architectures. We have used them during the development and evaluation of a search method for tuning space proposed in [1]. With our framework Kernel Tuning Toolkit, freely available at Github, we measured computation times and hardware performance counters on several GPUs for the complete tuning spaces of five benchmarks. These data, which we provide here, might benefit research of search algorithms for the tuning spaces of GPU codes or research of relation between applied code optimization, hardware performance counters, and GPU kernels’ performance.Moreover, we describe the scripts we used for robust evaluation of our searcher and comparison to others in detail. In particular, the script that simulates the tuning, i.e., replaces time-demanding compiling and executing the tuned kernels with a quick reading of the computation time from our measured data, makes it possible to inspect the convergence of tuning search over a large number of experiments. These scripts, freely available with our other codes, make it easier to experiment with search algorithms and compare them in a robust and reproducible way.During our research, we generated models for predicting values of performance counters from values of tuning parameters of our benchmarks. Here, we provide the models themselves and describe the scripts we implemented for their training. These data might benefit researchers who want to reproduce or build on our research.Jana HozzováJiří FilipovičAmin NezaratJaroslav Ol’haFilip PetrovičElsevierarticleAuto-tuningTuning spacesPerformance countersCudaComputer applications to medicine. Medical informaticsR858-859.7Science (General)Q1-390ENData in Brief, Vol 39, Iss , Pp 107631- (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Auto-tuning Tuning spaces Performance counters Cuda Computer applications to medicine. Medical informatics R858-859.7 Science (General) Q1-390
spellingShingle	Auto-tuning Tuning spaces Performance counters Cuda Computer applications to medicine. Medical informatics R858-859.7 Science (General) Q1-390 Jana Hozzová Jiří Filipovič Amin Nezarat Jaroslav Ol’ha Filip Petrovič Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
description	We have developed several autotuning benchmarks in CUDA that take into account performance-relevant source-code parameters and reach near peak-performance on various GPU architectures. We have used them during the development and evaluation of a search method for tuning space proposed in [1]. With our framework Kernel Tuning Toolkit, freely available at Github, we measured computation times and hardware performance counters on several GPUs for the complete tuning spaces of five benchmarks. These data, which we provide here, might benefit research of search algorithms for the tuning spaces of GPU codes or research of relation between applied code optimization, hardware performance counters, and GPU kernels’ performance.Moreover, we describe the scripts we used for robust evaluation of our searcher and comparison to others in detail. In particular, the script that simulates the tuning, i.e., replaces time-demanding compiling and executing the tuned kernels with a quick reading of the computation time from our measured data, makes it possible to inspect the convergence of tuning search over a large number of experiments. These scripts, freely available with our other codes, make it easier to experiment with search algorithms and compare them in a robust and reproducible way.During our research, we generated models for predicting values of performance counters from values of tuning parameters of our benchmarks. Here, we provide the models themselves and describe the scripts we implemented for their training. These data might benefit researchers who want to reproduce or build on our research.
format	article
author	Jana Hozzová Jiří Filipovič Amin Nezarat Jaroslav Ol’ha Filip Petrovič
author_facet	Jana Hozzová Jiří Filipovič Amin Nezarat Jaroslav Ol’ha Filip Petrovič
author_sort	Jana Hozzová
title	Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
title_short	Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
title_full	Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
title_fullStr	Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
title_full_unstemmed	Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures
title_sort	searching cuda code autotuning spaces with hardware performance counters: data from benchmarks running on various gpu architectures
publisher	Elsevier
publishDate	2021
url	https://doaj.org/article/18d5ea99c4e041b1adf14cc8b3d5c466
work_keys_str_mv	AT janahozzova searchingcudacodeautotuningspaceswithhardwareperformancecountersdatafrombenchmarksrunningonvariousgpuarchitectures AT jirifilipovic searchingcudacodeautotuningspaceswithhardwareperformancecountersdatafrombenchmarksrunningonvariousgpuarchitectures AT aminnezarat searchingcudacodeautotuningspaceswithhardwareperformancecountersdatafrombenchmarksrunningonvariousgpuarchitectures AT jaroslavolha searchingcudacodeautotuningspaceswithhardwareperformancecountersdatafrombenchmarksrunningonvariousgpuarchitectures AT filippetrovic searchingcudacodeautotuningspaceswithhardwareperformancecountersdatafrombenchmarksrunningonvariousgpuarchitectures
_version_	1718408303228223488

Searching CUDA code autotuning spaces with hardware performance counters: data from benchmarks running on various GPU architectures

Ejemplares similares