An OpenCL runtime system for a heterogeneous many-core virtual platform

Author

Kuan-Chung Chen ; Chung-Ho Chen

Author_Institution

Inst. of Comput. & Commun. Eng., Nat. Cheng Kung Univ., Tainan, Taiwan

fYear

2014

fDate

1-5 June 2014

Firstpage

2197

Lastpage

2200

Abstract

We present a many-core full system simulation platform and its OpenCL runtime system. The OpenCL runtime system includes an on-the-fly compiler and resource manager for the ARM-based many-core platform. Using this platform, we evaluate approaches of work-item scheduling and memory management in OpenCL memory hierarchy. Our experimental results show that scheduling work-items on a many-core system using general purpose RISC CPU should avoid per work-item context switching. Data deployment and work-item coalescing are the two keys for significant speedup.

Keywords

microcontrollers; multiprocessing systems; parallel programming; processor scheduling; program compilers; storage management; ARM-based many-core platform; OpenCL memory hierarchy; OpenCL runtime system; data deployment; general purpose RISC CPU; heterogeneous many-core virtual platform; many-core full system simulation platform; memory management; on-the-fly compiler; per work-item context switching; resource manager; work-item coalescing; work-item scheduling; Computational modeling; Instruction sets; Kernel; Memory management; Runtime; Switches; OpenCL; full system simulation; heterogeneous integration; runtime system; work-item coalescing;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems (ISCAS), 2014 IEEE International Symposium on

Conference_Location

Melbourne VIC

Print_ISBN

978-1-4799-3431-7

Type

conf

DOI

10.1109/ISCAS.2014.6865605

Filename

6865605