OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter

Pierre Olivier; Sang-hoon Kim; Binoy Ravindran

doi:10.1145/3102980.3103009

OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter

Pierre Olivier, Sang-hoon Kim, Binoy Ravindran

Advanced Processor Technology

Virginia Tech

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

The datacenter is becoming fully heterogeneous, integrating multiple OS-capable CPUs of different Instruction Set Architectures in separate machines. These machines present diverse performance and power consumption profiles and we show that significant potential benefits for both metrics can be expected, should these machines be able to cooperate in the processing of datacenter, multi-programmed workloads. We advocate that this cooperation should be enabled at the level of the OS, relieving the programmer from any effort related to the heterogeneity of the managed machines. We propose a distributed OS architecture running on a fully heterogeneous computer cluster, enabling this cooperation through three main components: the abstraction of the entire cluster in a single system image, a distributed shared memory system, and a heterogeneous scheduler.

Original language	English
Title of host publication	HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems
Publisher	Association for Computing Machinery
Pages	174-179
ISBN (Print)	978-1-4503-5068-6
DOIs	https://doi.org/10.1145/3102980.3103009
Publication status	Published - 1 May 2017
Event	HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems - Whistler, Canada Duration: 7 May 2017 → 10 May 2017

Conference

Conference	HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems
Country/Territory	Canada
City	Whistler
Period	7/05/17 → 10/05/17

Access to Document

10.1145/3102980.3103009

http://dl.acm.org/citation.cfm?doid=3102980.3103009

Cite this

@inproceedings{3ecd183499504f7fad4135507fdd472b,

title = "OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter",

abstract = "The datacenter is becoming fully heterogeneous, integrating multiple OS-capable CPUs of different Instruction Set Architectures in separate machines. These machines present diverse performance and power consumption profiles and we show that significant potential benefits for both metrics can be expected, should these machines be able to cooperate in the processing of datacenter, multi-programmed workloads. We advocate that this cooperation should be enabled at the level of the OS, relieving the programmer from any effort related to the heterogeneity of the managed machines. We propose a distributed OS architecture running on a fully heterogeneous computer cluster, enabling this cooperation through three main components: the abstraction of the entire cluster in a single system image, a distributed shared memory system, and a heterogeneous scheduler.",

author = "Pierre Olivier and Sang-hoon Kim and Binoy Ravindran",

year = "2017",

month = may,

day = "1",

doi = "10.1145/3102980.3103009",

language = "English",

isbn = "978-1-4503-5068-6",

pages = "174--179",

booktitle = "HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems",

publisher = "Association for Computing Machinery",

address = "United States",

note = "HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems ; Conference date: 07-05-2017 Through 10-05-2017",

}

Olivier, P, Kim, S & Ravindran, B 2017, OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter. in HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems. Association for Computing Machinery, pp. 174-179, HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems, Whistler, British Columbia, Canada, 7/05/17. https://doi.org/10.1145/3102980.3103009

OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter. / Olivier, Pierre; Kim, Sang-hoon; Ravindran, Binoy.

HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems. Association for Computing Machinery, 2017. p. 174-179.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter

AU - Olivier, Pierre

AU - Kim, Sang-hoon

AU - Ravindran, Binoy

PY - 2017/5/1

Y1 - 2017/5/1

N2 - The datacenter is becoming fully heterogeneous, integrating multiple OS-capable CPUs of different Instruction Set Architectures in separate machines. These machines present diverse performance and power consumption profiles and we show that significant potential benefits for both metrics can be expected, should these machines be able to cooperate in the processing of datacenter, multi-programmed workloads. We advocate that this cooperation should be enabled at the level of the OS, relieving the programmer from any effort related to the heterogeneity of the managed machines. We propose a distributed OS architecture running on a fully heterogeneous computer cluster, enabling this cooperation through three main components: the abstraction of the entire cluster in a single system image, a distributed shared memory system, and a heterogeneous scheduler.

AB - The datacenter is becoming fully heterogeneous, integrating multiple OS-capable CPUs of different Instruction Set Architectures in separate machines. These machines present diverse performance and power consumption profiles and we show that significant potential benefits for both metrics can be expected, should these machines be able to cooperate in the processing of datacenter, multi-programmed workloads. We advocate that this cooperation should be enabled at the level of the OS, relieving the programmer from any effort related to the heterogeneity of the managed machines. We propose a distributed OS architecture running on a fully heterogeneous computer cluster, enabling this cooperation through three main components: the abstraction of the entire cluster in a single system image, a distributed shared memory system, and a heterogeneous scheduler.

U2 - 10.1145/3102980.3103009

DO - 10.1145/3102980.3103009

M3 - Conference contribution

SN - 978-1-4503-5068-6

SP - 174

EP - 179

BT - HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems

PB - Association for Computing Machinery

T2 - HotOS '17: Proceedings of the 16th Workshop on Hot Topics in Operating Systems

Y2 - 7 May 2017 through 10 May 2017

ER -

OS Support for Thread Migration and Distribution in the Fully Heterogeneous Datacenter

Abstract

Conference

Access to Document

Fingerprint

Cite this