DMCA
OP2: An Active Library Framework for Solving Unstructured Mesh-based Applications on Multi-Core and Many-Core Architectures ∗
Cached
Download Links
| Citations: | 9 - 3 self |
BibTeX
@MISC{Mudalige_op2:an,
author = {G. R. Mudalige and M. B. Giles and I. Reguly and C. Bertolli and P. H. J Kelly},
title = {OP2: An Active Library Framework for Solving Unstructured Mesh-based Applications on Multi-Core and Many-Core Architectures ∗},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
OP2 is an “active ” library framework for the solution of unstructured mesh-based applications. It utilizes sourceto-source translation and compilation so that a single application code written using the OP2 API can be transformed into different parallel implementations for execution on different back-end hardware platforms. In this paper we present the design of the current OP2 library, and investigate its capabilities in achieving performance portability, near-optimal performance, and scaling on modern multi-core and many-core processor based systems. A key feature of this work is OP2’s recent extension facilitating the development and execution of applications on a distributed memory cluster of GPUs. We discuss the main design issues in parallelizing unstructured mesh based applications on heterogeneous platforms. These include handling data dependencies in accessing indirectly referenced data, the impact of unstructured mesh data layouts (array of structs vs. struct of arrays) and design considerations in generating code for execution on a cluster of GPUs. A representative CFD application written using the OP2 framework is utilized to provide a contrasting benchmarking and performance analysis study on a range of multicore/many-core systems. These include multi-core CPUs
Keyphrases
active library framework unstructured mesh-based application many-core architecture unstructured mesh multi-core cpu modern multi-core near-optimal performance unstructured mesh data layout key feature contrasting benchmarking op2 recent extension design consideration op2 framework current op2 library main design issue sourceto-source translation distributed memory cluster different back-end hardware platform performance portability multicore many-core system data dependency different parallel implementation many-core processor single application code heterogeneous platform op2 api representative cfd application performance analysis study
