Power, Reliability, and Performance: One System to Rule them All

Bilge Acun; Akhil Langer; Esteban Meneses; Harshitha Menon; Osman Sarood; Ehsan Totoni; Laxmikant V. Kalé

doi:10.1109/MC.2016.310

Power, Reliability, and Performance: One System to Rule them All

Bilge Acun, Akhil Langer, Esteban Meneses, Harshitha Menon, Osman Sarood, Ehsan Totoni, Laxmikant V. Kalé

Research output: Contribution to specialist publication › Article

16 Scopus citations

Abstract

In a design based on the Charm++ parallel programming framework, an adaptive runtime system dynamically interacts with a datacenter's resource manager to control power by intelligently scheduling jobs, reallocating resources, and reconfiguring hardware. It simultaneously manages reliability by cooling the system to the running application's optimal level and maintains performance through load balancing.

Original language	English
Pages	30-37
Number of pages	8
Volume	49
No	10
Specialist publication	Computer
DOIs	https://doi.org/10.1109/MC.2016.310
State	Published - Oct 2016

Keywords

Charm++
energy-aware systems
energy-efficient computing
high-performance computing
HPC
parallel systems
power management
temperature-aware design

Access to Document

10.1109/MC.2016.310

Cite this

@misc{8d8b8a78541746d890febc0711fccab8,

title = "Power, Reliability, and Performance: One System to Rule them All",

abstract = "In a design based on the Charm++ parallel programming framework, an adaptive runtime system dynamically interacts with a datacenter's resource manager to control power by intelligently scheduling jobs, reallocating resources, and reconfiguring hardware. It simultaneously manages reliability by cooling the system to the running application's optimal level and maintains performance through load balancing.",

keywords = "Charm++, energy-aware systems, energy-efficient computing, high-performance computing, HPC, parallel systems, power management, temperature-aware design",

author = "Bilge Acun and Akhil Langer and Esteban Meneses and Harshitha Menon and Osman Sarood and Ehsan Totoni and Kal{\'e}, {Laxmikant V.}",

note = "Publisher Copyright: {\textcopyright} 1970-2012 IEEE.",

year = "2016",

month = oct,

doi = "10.1109/MC.2016.310",

language = "Ingl{\'e}s",

volume = "49",

pages = "30--37",

journal = "Computer",

issn = "0018-9162",

}

TY - GEN

T1 - Power, Reliability, and Performance

T2 - One System to Rule them All

AU - Acun, Bilge

AU - Langer, Akhil

AU - Meneses, Esteban

AU - Menon, Harshitha

AU - Sarood, Osman

AU - Totoni, Ehsan

AU - Kalé, Laxmikant V.

PY - 2016/10

Y1 - 2016/10

N2 - In a design based on the Charm++ parallel programming framework, an adaptive runtime system dynamically interacts with a datacenter's resource manager to control power by intelligently scheduling jobs, reallocating resources, and reconfiguring hardware. It simultaneously manages reliability by cooling the system to the running application's optimal level and maintains performance through load balancing.

AB - In a design based on the Charm++ parallel programming framework, an adaptive runtime system dynamically interacts with a datacenter's resource manager to control power by intelligently scheduling jobs, reallocating resources, and reconfiguring hardware. It simultaneously manages reliability by cooling the system to the running application's optimal level and maintains performance through load balancing.

KW - Charm++

KW - energy-aware systems

KW - energy-efficient computing

KW - high-performance computing

KW - HPC

KW - parallel systems

KW - power management

KW - temperature-aware design

UR - http://www.scopus.com/inward/record.url?scp=84992741049&partnerID=8YFLogxK

U2 - 10.1109/MC.2016.310

DO - 10.1109/MC.2016.310

M3 - Artículo

AN - SCOPUS:84992741049

SN - 0018-9162

VL - 49

SP - 30

EP - 37

JO - Computer

JF - Computer

ER -

Power, Reliability, and Performance: One System to Rule them All

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this