Comparative study of high performance software rasterization techniques

Frolov, V.A.; Galaktionov, V.A.; Barladyan, B.H.

Авторы: Frolov V.A., Galaktionov V.A., Barladyan B.H.
Журнал: Mathematica Montisnigri
Том: 47
Год издания: 2020
Первая страница: 152
Последняя страница: 175
DOI: 10.20948/mathmontis-2020-47-13
Аннотация: This paper provides a comparative study and performance analysis ofdi erent rasterization algorithms and approaches. Unlike many other papers, we don'tfocus on rasterization itself, but investigate complete graphics pipeline with 3D transfor-mations, Z-bu er, perspective correction and texturing that, on the one hand, allow usto implement a useful subset of OpenGL functionality and, on the other hand, considervarious bottlenecks in the graphics pipeline and how di erent approaches manage them.Our ultimate goal is to nd a scalable rasterizer technique that on the one hand e ectivelyuses current CPUs and on the other hand is accelerating with the extensive developmentof hardware. We explore the capabilities of scan-line and half-space algorithms raster-ization, investigate di erent memory layout for frame bu er data, study the possibilityof instruction-level and thread-level parallelism to be applied. We also study relativeeciency of di erent CPU architectures (in-order CPUs vs out-of-order CPUs) for thegraphics pipeline implementation and tested our solution with x64, ARMv7 and ARMv8instruction sets. We were able to propose an approach that could outperform highly op-timized OpenSWR rasterizer for small triangles. Finally, we conclude that despite a hugebackground high-performance software rasterization still has a lot of interesting topics forfuture research.
Добавил в систему: Фролов Владимир Александрович

	ИСТИНА	Войти в систему Регистрация
	ИСТИНА ИНХС РАН
	Главная Поиск Статистика О проекте Помощь

ИСТИНА

ИСТИНА ИНХС РАН

Comparative study of high performance software rasterization techniquesстатья

Прикрепленные файлы