Portable and Transparent Software Managed Scheduling on Accelerators for Fair Resource Sharing

Christos Margiolas, Michael F. P. O'Boyle

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Accelerators, such as Graphic Processing Units (GPUs), are popular components of modern parallel systems. Their energy-efficient performance make them attractive components for modern data center nodes. However, they lack control for fair resource sharing amongst multiple users. This paper presents a runtime and Just In Time compiler that enables resource sharing control and software managed scheduling on accelerators. It is portable and transparent, requiring no modification or recompilation of existing systems or user applications. We provide an extensive evaluation of our scheme with over 40,000 different workloads on 2 platforms and we deliver fairness improvements ranging from 6.8x to 13.66x. In addition, we also deliver system throughput speedups ranging from 1.13x to 1.31x.
Original languageEnglish
Title of host publicationCGO 2016 Proceedings of the 2016 International Symposium on Code Generation and Optimization
PublisherACM
Pages82-93
Number of pages12
ISBN (Print)978-1-4503-3778-6
DOIs
Publication statusPublished - 2016
Event2016 International Symposium on Code Generation and Optimization - Barcelona, Spain
Duration: 12 Mar 201618 Mar 2016
http://cgo.org/cgo2016/

Conference

Conference2016 International Symposium on Code Generation and Optimization
Abbreviated titleCGO 2016
CountrySpain
CityBarcelona
Period12/03/1618/03/16
Internet address

Fingerprint Dive into the research topics of 'Portable and Transparent Software Managed Scheduling on Accelerators for Fair Resource Sharing'. Together they form a unique fingerprint.

Cite this