Automatic co-scheduling based on main memory bandwidth usage

Jens Breitbart, Josef Weidendorfer, Carsten Trinitis

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Most applications running on supercomputers achieve only a fraction of a system’s peak performance. It has been demonstrated that co-scheduling applications can improve overall system utilization. In this case, however, applications being co-scheduled need to fulfill certain criteria such that mutual slowdown is kept at a minimum. In this paper we present a set of libraries and a first HPC scheduler prototype that automatically detects an application’s main memory bandwidth utilization and prevents the co-scheduling of multiple main memory bandwidth limited applications. We demonstrate that our prototype achieves almost the same performance as we achieved with manually tuned co-schedules in previous work.

Original languageEnglish
Title of host publicationJob Scheduling Strategies for Parallel Processing - 19th and 20th International Workshops, JSSPP 2015, Revised Selected Papers
EditorsNarayan Desai, Walfredo Cirne
PublisherSpringer Verlag
Pages141-157
Number of pages17
ISBN (Print)9783319617558
DOIs
StatePublished - 2017
Event20th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2016 - Chicago, United States
Duration: 27 May 201627 May 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10353 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2016
Country/TerritoryUnited States
CityChicago
Period27/05/1627/05/16

Fingerprint

Dive into the research topics of 'Automatic co-scheduling based on main memory bandwidth usage'. Together they form a unique fingerprint.

Cite this