|
ИСТИНА |
Войти в систему Регистрация |
ИСТИНА ИНХС РАН |
||
An Efficient Implementation of the Transitive Closure Problem on Intel KNL Architectureстатья
- Автор: Afanasyev Ilya
- Сборник: Proceedings of the 3rd Ural Workshop on Parallel, Distributed, and Cloud Computing for Young Scientists
- Год издания: 2017
- Первая страница: 10
- Последняя страница: 19
- Аннотация: An important trend in modern supercomputing is a frequent usage of co-processors, such as GPUs and Intel Xeon PHIs. The recent generation of Intel Knights Landing processors provide high performance computational power with a large amount of high-bandwidth memory, what makes them a perfect platform for graph-processing. The presented study describes implementation approaches to large-scale graph process- ing on Intel KNL processors; as a sample problem, the transitive closure computation is discussed. Based on the joint analysis of algorithm prop- erties and architecture features, the performance tuning has been per- formed, including graph storage format optimizations, efficient usage of memory hierarchy and vectorization. As a result, an optimized algorithm implementation for the transitive closure problem solution has been de- veloped. The proposed implementation has been studied using different approaches, aimed at demonstrating advantages and disadvantages of Intel KNL architecture in solving graph-processing problems.
- Добавил в систему: Афанасьев Илья Викторович