Hybrid Embarrassingly Parallel on heterogeneous platform

Author

Gong, Chunye ; Liu, Jie ; Qin, Jin ; Hu, Qingfeng ; Gong, Zhenghu

Author_Institution

Dept. of Comput. Sci., Nat. Univ. of Defense Technol., Changsha, China

Volume

3

fYear

2010

fDate

9-11 July 2010

Firstpage

95

Lastpage

99

Abstract

The Embarrassingly Parallel (EP) is one kernel benchmark of NAS Parallel Benchmarks (NPB). EP generates pairs of Gaussian Random Deviates (GRDs) of large random numbers which produced by Linear Congruential Generator (LCG). In this paper, the Hybrid EP is efficient implemented on CPU/GPU heterogeneous platform. Experimental results show that the Hybrid EP is 11.98 times faster than equivalent multicore CPU and outperforms equivalent GPU by 7.52%.

Keywords

benchmark testing; computer graphic equipment; coprocessors; parallel architectures; random number generation; CPU-GPU heterogeneous platform; Gaussian random deviation; NAS parallel benchmark; equivalent multicore CPU; hybrid EP; hybrid embarrassingly parallel; kernel benchmark; linear congruential generator; random numbers; Acceleration; Europe; Graphics processing unit; Laboratories; Optimization; GPU; Hybrid EP; NPB Embarrassingly Parallel (EP); heterogeneous platform;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Information Technology (ICCSIT), 2010 3rd IEEE International Conference on

Conference_Location

Chengdu

Print_ISBN

978-1-4244-5537-9

Type

conf

DOI

10.1109/ICCSIT.2010.5564662

Filename

5564662