The increasing complexity of modern and future computing systems makes it challenging to develop applications that aim for maximum performance. Hybrid parallelprogrammingmodels offer new ways to exploit the capabili...
详细信息
ISBN:
(纸本)9781728165820
The increasing complexity of modern and future computing systems makes it challenging to develop applications that aim for maximum performance. Hybrid parallelprogrammingmodels offer new ways to exploit the capabilities of the underlying infrastructure. However, the performance gain is sometimes accompanied by increased programming complexity. We introduce an extension to PyCOMPSs, a high-level task-basedparallelprogramming model for Python applications, to support tasks that use MPI natively as part of the task model. Without compromising application's programmability, using Native MPI tasks in PyCOMPSs offers up to 3x improvement in total performance for compute intensive applications and up to 1.9x improvement in total performance for 110 intensive applications over sequential implementation of the tasks.
Distributed computing platforms are evolving to heterogeneous ecosystems with Clusters, Grids and Clouds introducing in its computing nodes, processors with different core architectures, accelerators (i.e. GPUs, FPGAs...
详细信息
Distributed computing platforms are evolving to heterogeneous ecosystems with Clusters, Grids and Clouds introducing in its computing nodes, processors with different core architectures, accelerators (i.e. GPUs, FPGAs), as well as different memories and storage devices in order to achieve better performance with lower energy consumption. As a consequence of this heterogeneity, programming applications for these distributed heterogeneous platforms becomes a complex task. Additionally to the complexity of developing an application for distributed platforms, developers must also deal now with the complexity of the different computing devices inside the node. In this article, we present a programming model that aims to facilitate the development and execution of applications in current and future distributed heterogeneous parallel architectures. This programming model is based on the hierarchical composition of the COMP Superscalar and Omp Superscalar programmingmodels that allow developers to implement infrastructure-agnostic applications. The underlying runtime enables applications to adapt to the infrastructure without the need of maintaining different versions of the code. Our programming model proposal has been evaluated on real platforms, in terms of heterogeneous resource usage, performance and adaptation.
暂无评论