期刊
JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING
卷 73, 期 11, 页码 1439-1450出版社
ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.jpdc.2012.07.005
关键词
Many-core computing; GPU computing; Optimisation; OpenCL; High performance computing
资金
- Royal Society [IF090020/AM]
- UK Atomic Weapons Establishment [CDK0660, CDK0724]
- AWE and the TSB Knowledge Transfer Partnership [KTP006740]
- United States Department of Energy's National Nuclear Security Administration [DE-AC04-94AL85000]
- Royal Society [IF090020] Funding Source: Royal Society
This paper reports on the development of an MPI/OpenCL implementation of LU, an application-level benchmark from the NAS Parallel Benchmark Suite. An account of the design decisions addressed during the development of this code is presented, demonstrating the importance of memory arrangement and work-item/work-group distribution strategies when applications are deployed on different device types. The resulting platform-agnostic, single source application is benchmarked on a number of different architectures, and is shown to be 1.3-1.5x slower than native FORTRAN 77 or CUDA implementations on a single node and 1.3-3.1x slower on multiple nodes. We also explore the potential performance gains of OpenCL's device fissioning capability, demonstrating up to a 3 x speed-up over our original OpenCL implementation. (C) 2012 Elsevier Inc. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据