Parallel density-based stream clustering using a multi-user GPU scheduler

A. Tarakji; M. Hassani; L. Georgiev; T. Seidl; R. Leupers

doi:10.1007/978-3-319-18422-7_31

Parallel density-based stream clustering using a multi-user GPU scheduler

A. Tarakji, M. Hassani, L. Georgiev, T. Seidl, R. Leupers

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

Abstract

With the emergence of advanced stream computing architectures, their deployment to accelerate long-running data mining applications is becoming a matter of course. This work presents a novel design concept of the stream clustering algorithm DenStream, based on a previously presented scheduling framework for GPUs. By means of our scheduler OCLSched, DenStream runs together with general computation tasks in a multi-user computing environment, sharing the GPU resources. A major point of concern throughout this paper has been to disclose the functionality and purposes of the applied scheduling methods, and to demonstrate the OCLSched’s ability of managing highly complex applications in a multi-task GPU environment. Also in terms of performance, our tests show reasonable improvements when comparing the proposed parallel concept of DenStream with a single-threaded CPU version

Original language	English
Title of host publication	Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings
Editors	S. Kozielski, D. Mrozek, P. Kasprowski, B. Małysiak-Mrozek, D. Kostrzewa
Place of Publication	Cham
Publisher	Springer
Pages	343-360
Number of pages	18
ISBN (Electronic)	978-3-319-18422-7
ISBN (Print)	978-3-319-18421-0
DOIs	https://doi.org/10.1007/978-3-319-18422-7_31
Publication status	Published - 26 May 2015
Externally published	Yes
Event	11th International Conference on Beyond Databases, Architectures and Structures (BDAS 2015) - Ustro ́n, Poland Duration: 26 May 2015 → 29 May 2015 Conference number: 11

Publication series

Name	Communications in Computer and Information Science
Volume	521
ISSN (Print)	1865-0929

Conference

Conference	11th International Conference on Beyond Databases, Architectures and Structures (BDAS 2015)
Abbreviated title	BDAS 2015
Country/Territory	Poland
City	Ustro ́n
Period	26/05/15 → 29/05/15

Keywords

GPGPU
OpenCL
Data mining
Streams
Task Scheduling

Access to Document

10.1007/978-3-319-18422-7_31

http://dx.doi.org/10.1007/978-3-319-18422-7_31

Cite this

Tarakji, A., Hassani, M., Georgiev, L., Seidl, T., & Leupers, R. (2015). Parallel density-based stream clustering using a multi-user GPU scheduler. In S. Kozielski, D. Mrozek, P. Kasprowski, B. Małysiak-Mrozek, & D. Kostrzewa (Eds.), Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings (pp. 343-360). (Communications in Computer and Information Science; Vol. 521). Springer. https://doi.org/10.1007/978-3-319-18422-7_31

Tarakji, A. ; Hassani, M. ; Georgiev, L. et al. / Parallel density-based stream clustering using a multi-user GPU scheduler. Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings. editor / S. Kozielski ; D. Mrozek ; P. Kasprowski ; B. Małysiak-Mrozek ; D. Kostrzewa. Cham : Springer, 2015. pp. 343-360 (Communications in Computer and Information Science).

@inproceedings{98c084d39eea4c1f82b7357ebb0e30b2,

title = "Parallel density-based stream clustering using a multi-user GPU scheduler",

abstract = "With the emergence of advanced stream computing architectures, their deployment to accelerate long-running data mining applications is becoming a matter of course. This work presents a novel design concept of the stream clustering algorithm DenStream, based on a previously presented scheduling framework for GPUs. By means of our scheduler OCLSched, DenStream runs together with general computation tasks in a multi-user computing environment, sharing the GPU resources. A major point of concern throughout this paper has been to disclose the functionality and purposes of the applied scheduling methods, and to demonstrate the OCLSched{\textquoteright}s ability of managing highly complex applications in a multi-task GPU environment. Also in terms of performance, our tests show reasonable improvements when comparing the proposed parallel concept of DenStream with a single-threaded CPU version",

keywords = "GPGPU, OpenCL, Data mining, Streams, Task Scheduling",

author = "A. Tarakji and M. Hassani and L. Georgiev and T. Seidl and R. Leupers",

year = "2015",

month = may,

day = "26",

doi = "10.1007/978-3-319-18422-7_31",

language = "English",

isbn = "978-3-319-18421-0",

series = "Communications in Computer and Information Science",

publisher = "Springer",

pages = "343--360",

editor = "S. Kozielski and D. Mrozek and P. Kasprowski and { Ma{\l}ysiak-Mrozek}, B. and D. Kostrzewa",

booktitle = "Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings",

address = "Germany",

note = "11th International Conference on Beyond Databases, Architectures and Structures (BDAS 2015), BDAS 2015 ; Conference date: 26-05-2015 Through 29-05-2015",

}

Tarakji, A, Hassani, M, Georgiev, L, Seidl, T & Leupers, R 2015, Parallel density-based stream clustering using a multi-user GPU scheduler. in S Kozielski, D Mrozek, P Kasprowski, B Małysiak-Mrozek & D Kostrzewa (eds), Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings. Communications in Computer and Information Science, vol. 521, Springer, Cham, pp. 343-360, 11th International Conference on Beyond Databases, Architectures and Structures (BDAS 2015), Ustro ́n, Poland, 26/05/15. https://doi.org/10.1007/978-3-319-18422-7_31

Parallel density-based stream clustering using a multi-user GPU scheduler. / Tarakji, A.; Hassani, M.; Georgiev, L. et al.

Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings. ed. / S. Kozielski; D. Mrozek; P. Kasprowski; B. Małysiak-Mrozek; D. Kostrzewa. Cham : Springer, 2015. p. 343-360 (Communications in Computer and Information Science; Vol. 521).

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

TY - GEN

T1 - Parallel density-based stream clustering using a multi-user GPU scheduler

AU - Tarakji, A.

AU - Hassani, M.

AU - Georgiev, L.

AU - Seidl, T.

AU - Leupers, R.

N1 - Conference code: 11

PY - 2015/5/26

Y1 - 2015/5/26

N2 - With the emergence of advanced stream computing architectures, their deployment to accelerate long-running data mining applications is becoming a matter of course. This work presents a novel design concept of the stream clustering algorithm DenStream, based on a previously presented scheduling framework for GPUs. By means of our scheduler OCLSched, DenStream runs together with general computation tasks in a multi-user computing environment, sharing the GPU resources. A major point of concern throughout this paper has been to disclose the functionality and purposes of the applied scheduling methods, and to demonstrate the OCLSched’s ability of managing highly complex applications in a multi-task GPU environment. Also in terms of performance, our tests show reasonable improvements when comparing the proposed parallel concept of DenStream with a single-threaded CPU version

AB - With the emergence of advanced stream computing architectures, their deployment to accelerate long-running data mining applications is becoming a matter of course. This work presents a novel design concept of the stream clustering algorithm DenStream, based on a previously presented scheduling framework for GPUs. By means of our scheduler OCLSched, DenStream runs together with general computation tasks in a multi-user computing environment, sharing the GPU resources. A major point of concern throughout this paper has been to disclose the functionality and purposes of the applied scheduling methods, and to demonstrate the OCLSched’s ability of managing highly complex applications in a multi-task GPU environment. Also in terms of performance, our tests show reasonable improvements when comparing the proposed parallel concept of DenStream with a single-threaded CPU version

KW - GPGPU

KW - OpenCL

KW - Data mining

KW - Streams

KW - Task Scheduling

U2 - 10.1007/978-3-319-18422-7_31

DO - 10.1007/978-3-319-18422-7_31

M3 - Conference contribution

SN - 978-3-319-18421-0

T3 - Communications in Computer and Information Science

SP - 343

EP - 360

BT - Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings

A2 - Kozielski, S.

A2 - Mrozek, D.

A2 - Kasprowski, P.

A2 - Małysiak-Mrozek, B.

A2 - Kostrzewa, D.

PB - Springer

CY - Cham

T2 - 11th International Conference on Beyond Databases, Architectures and Structures (BDAS 2015)

Y2 - 26 May 2015 through 29 May 2015

ER -

Tarakji A, Hassani M, Georgiev L, Seidl T, Leupers R. Parallel density-based stream clustering using a multi-user GPU scheduler. In Kozielski S, Mrozek D, Kasprowski P, Małysiak-Mrozek B, Kostrzewa D, editors, Beyond Databases, Architectures and Structures - 11th International Conference, BDAS 2015, Ustron, Poland, May 26-29, 2015, Proceedings. Cham: Springer. 2015. p. 343-360. (Communications in Computer and Information Science). https://doi.org/10.1007/978-3-319-18422-7_31

Parallel density-based stream clustering using a multi-user GPU scheduler

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this