Value and energy optimizing dynamic resource allocation in many-core HPC systems

TY - GEN

T1 - Value and energy optimizing dynamic resource allocation in many-core HPC systems

AU - Singh, Amit Kumar

AU - Dziurzanski, Piotr

AU - Indrusiak, Leandro Soares

N1 - © IEEE, 2016. This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

PY - 2016/2/1

Y1 - 2016/2/1

N2 - The conventional approaches to reduce the energy consumption of high performance computing (HPC) data centers focus on consolidation and dynamic voltage and frequency scaling (DVFS). Most of these approaches consider independent tasks (or jobs) and do not jointly optimize for energy and value. In this paper, we propose DVFS-aware profiling and non-profiling based approaches that use design-time profiling results and perform all the computations at run-time, respectively. The profiling based approach is suitable for the scenarios when the jobs or their structure is known at design-time, otherwise, the non-profiling based approach is more suitable. Both the approaches consider jobs containing dependent tasks and exploit efficient allocation combined with identification of voltage/frequency levels of used system cores to jointly optimize value and energy. Experiments show that the proposed approaches reduce energy consumption by 15% when compared to existing approaches while achieving significant amount of value and reducing percentage of rejected jobs leading to zero value.

AB - The conventional approaches to reduce the energy consumption of high performance computing (HPC) data centers focus on consolidation and dynamic voltage and frequency scaling (DVFS). Most of these approaches consider independent tasks (or jobs) and do not jointly optimize for energy and value. In this paper, we propose DVFS-aware profiling and non-profiling based approaches that use design-time profiling results and perform all the computations at run-time, respectively. The profiling based approach is suitable for the scenarios when the jobs or their structure is known at design-time, otherwise, the non-profiling based approach is more suitable. Both the approaches consider jobs containing dependent tasks and exploit efficient allocation combined with identification of voltage/frequency levels of used system cores to jointly optimize value and energy. Experiments show that the proposed approaches reduce energy consumption by 15% when compared to existing approaches while achieving significant amount of value and reducing percentage of rejected jobs leading to zero value.

KW - Energy consumption

KW - High Performance Computing

KW - Many-core

KW - Resource allocation

KW - Value

KW - Value curves

UR - http://www.scopus.com/inward/record.url?scp=84964378681&partnerID=8YFLogxK

U2 - 10.1109/CloudCom.2015.22

DO - 10.1109/CloudCom.2015.22

M3 - Conference contribution

AN - SCOPUS:84964378681

SN - 9781467395601

SP - 180

EP - 185

BT - 2015 IEEE 7th International Conference on Cloud Computing Technology and Science (CloudCom)

PB - IEEE

T2 - 7th IEEE International Conference on Cloud Computing Technology and Science, CloudCom 2015

Y2 - 30 November 2015 through 3 December 2015

ER -