PACUE: Processor allocator considering user experience

Tetsuro Horikawa; Michio Honda; Jin Nakazawa; Kazunori Takashio; Hideyuki Tokuda

doi:10.1007/978-3-642-29740-3_38

PACUE: Processor allocator considering user experience

Tetsuro Horikawa, Michio Honda, Jin Nakazawa, Kazunori Takashio, Hideyuki Tokuda

Faculty of Environment and Information Studies

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

1 Citation (Scopus)

Abstract

GPU accelerated applications including GPGPU ones are commonly seen in modern PCs. If many applications compete on the same GPU, the performance will decrease significantly. Some applications have a large impact on user experience. Therefore, for such applications, we have to limit GPU utilization by the other applications. It might be straightforward to modify applications to switch compute device dynamically for intelligent resources allocation. Unfortunately, we cannot do so due to software distribution policy or the other reasons. In this paper, we propose PACUE, which allows the end system to allocate compute devices arbitrary to applications. In addition, PACUE guesses optimal compute device for each application according to user preference. We implemented the dynamic compute device redirector of PACUE including OpenCL API hooking and device camouflaging features. We also implemented the frame of the resource manager of PACUE. We demonstrate PACUE achieves dynamic compute device redirecting on one out of two real applications and on all of 20 sample codes.

Original language	English
Title of host publication	Euro-Par 2011
Subtitle of host publication	Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers
Publisher	Springer Verlag
Pages	335-344
Number of pages	10
Edition	PART 2
ISBN (Print)	9783642297397
DOIs	https://doi.org/10.1007/978-3-642-29740-3_38
Publication status	Published - 2012
Event	17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011 - Bordeaux, France Duration: 2011 Aug 29 → 2011 Sept 2

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	7156 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011
Country/Territory	France
City	Bordeaux
Period	11/8/29 → 11/9/2

Keywords

GPGPU
GPU
OpenCL
PC
Resource management
binary compatibility
user experience

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-642-29740-3_38

Cite this

Horikawa, T., Honda, M., Nakazawa, J., Takashio, K., & Tokuda, H. (2012). PACUE: Processor allocator considering user experience. In Euro-Par 2011: Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers (PART 2 ed., pp. 335-344). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7156 LNCS, No. PART 2). Springer Verlag. https://doi.org/10.1007/978-3-642-29740-3_38

PACUE : Processor allocator considering user experience. / Horikawa, Tetsuro; Honda, Michio; Nakazawa, Jin et al.

Euro-Par 2011: Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers. PART 2. ed. Springer Verlag, 2012. p. 335-344 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7156 LNCS, No. PART 2).

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

Horikawa, T, Honda, M, Nakazawa, J , Takashio, K & Tokuda, H 2012, PACUE: Processor allocator considering user experience. in Euro-Par 2011: Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 7156 LNCS, Springer Verlag, pp. 335-344, 17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011, Bordeaux, France, 11/8/29. https://doi.org/10.1007/978-3-642-29740-3_38

Horikawa T, Honda M, Nakazawa J , Takashio K, Tokuda H. PACUE: Processor allocator considering user experience. In Euro-Par 2011: Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers. PART 2 ed. Springer Verlag. 2012. p. 335-344. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-642-29740-3_38

Horikawa, Tetsuro ; Honda, Michio ; Nakazawa, Jin et al. / PACUE : Processor allocator considering user experience. Euro-Par 2011: Parallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers. PART 2. ed. Springer Verlag, 2012. pp. 335-344 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{8f212b677d8740829d0625a9771f2037,

title = "PACUE: Processor allocator considering user experience",

abstract = "GPU accelerated applications including GPGPU ones are commonly seen in modern PCs. If many applications compete on the same GPU, the performance will decrease significantly. Some applications have a large impact on user experience. Therefore, for such applications, we have to limit GPU utilization by the other applications. It might be straightforward to modify applications to switch compute device dynamically for intelligent resources allocation. Unfortunately, we cannot do so due to software distribution policy or the other reasons. In this paper, we propose PACUE, which allows the end system to allocate compute devices arbitrary to applications. In addition, PACUE guesses optimal compute device for each application according to user preference. We implemented the dynamic compute device redirector of PACUE including OpenCL API hooking and device camouflaging features. We also implemented the frame of the resource manager of PACUE. We demonstrate PACUE achieves dynamic compute device redirecting on one out of two real applications and on all of 20 sample codes.",

keywords = "GPGPU, GPU, OpenCL, PC, Resource management, binary compatibility, user experience",

author = "Tetsuro Horikawa and Michio Honda and Jin Nakazawa and Kazunori Takashio and Hideyuki Tokuda",

note = "Copyright: Copyright 2021 Elsevier B.V., All rights reserved.; 17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011 ; Conference date: 29-08-2011 Through 02-09-2011",

year = "2012",

doi = "10.1007/978-3-642-29740-3_38",

language = "English",

isbn = "9783642297397",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

number = "PART 2",

pages = "335--344",

booktitle = "Euro-Par 2011",

edition = "PART 2",

}

TY - GEN

T1 - PACUE

T2 - 17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011

AU - Horikawa, Tetsuro

AU - Honda, Michio

AU - Nakazawa, Jin

AU - Takashio, Kazunori

AU - Tokuda, Hideyuki

PY - 2012

Y1 - 2012

N2 - GPU accelerated applications including GPGPU ones are commonly seen in modern PCs. If many applications compete on the same GPU, the performance will decrease significantly. Some applications have a large impact on user experience. Therefore, for such applications, we have to limit GPU utilization by the other applications. It might be straightforward to modify applications to switch compute device dynamically for intelligent resources allocation. Unfortunately, we cannot do so due to software distribution policy or the other reasons. In this paper, we propose PACUE, which allows the end system to allocate compute devices arbitrary to applications. In addition, PACUE guesses optimal compute device for each application according to user preference. We implemented the dynamic compute device redirector of PACUE including OpenCL API hooking and device camouflaging features. We also implemented the frame of the resource manager of PACUE. We demonstrate PACUE achieves dynamic compute device redirecting on one out of two real applications and on all of 20 sample codes.

AB - GPU accelerated applications including GPGPU ones are commonly seen in modern PCs. If many applications compete on the same GPU, the performance will decrease significantly. Some applications have a large impact on user experience. Therefore, for such applications, we have to limit GPU utilization by the other applications. It might be straightforward to modify applications to switch compute device dynamically for intelligent resources allocation. Unfortunately, we cannot do so due to software distribution policy or the other reasons. In this paper, we propose PACUE, which allows the end system to allocate compute devices arbitrary to applications. In addition, PACUE guesses optimal compute device for each application according to user preference. We implemented the dynamic compute device redirector of PACUE including OpenCL API hooking and device camouflaging features. We also implemented the frame of the resource manager of PACUE. We demonstrate PACUE achieves dynamic compute device redirecting on one out of two real applications and on all of 20 sample codes.

KW - GPGPU

KW - GPU

KW - OpenCL

KW - PC

KW - Resource management

KW - binary compatibility

KW - user experience

UR - http://www.scopus.com/inward/record.url?scp=84882628602&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84882628602&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-29740-3_38

DO - 10.1007/978-3-642-29740-3_38

M3 - Conference contribution

AN - SCOPUS:84882628602

SN - 9783642297397

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 335

EP - 344

BT - Euro-Par 2011

PB - Springer Verlag

Y2 - 29 August 2011 through 2 September 2011

ER -

PACUE: Processor allocator considering user experience

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this