Elastic Management of Reconfigurable Accelerators

Author

Grigoras, Paul ; Tottenham, Max ; Xinyu Niu ; Coutinho, J.G.F. ; Luk, Wayne

Author_Institution

Dept. of Comput., Imperial Coll. London, London, UK

fYear

2014

fDate

26-28 Aug. 2014

Firstpage

174

Lastpage

181

Abstract

This paper presents a runtime system for reconfigurable accelerators that supports elastic management: it enables effective sharing of accelerator resources across multiple applications. For each application, this runtime system allocates an appropriate amount of resources to satisfy its quality-of-service requirements, while minimising the overall execution time for a collection of applications. The effectiveness of this runtime system is due to a set of scheduling algorithms and strategies customised for different types of workloads. We demonstrate our approach by implementing a dynamic Monte Carlo bond options pricing design.

Keywords

cloud computing; field programmable gate arrays; reconfigurable architectures; resource allocation; scheduling; FPGA; accelerator resources sharing; cloud computing; dynamic Monte Carlo bond options pricing design; elastic management; field programmable gate array; quality-of-service requirements; reconfigurable accelerators; resources allocation; runtime system; scheduling algorithms; Dynamic scheduling; Elasticity; Field programmable gate arrays; Hydrogen; Resource management; Runtime; Scheduling algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel and Distributed Processing with Applications (ISPA), 2014 IEEE International Symposium on

Conference_Location

Milan

Type

conf

DOI

10.1109/ISPA.2014.31

Filename

6924444