Aspects of GPU for general purpose high performance computing

Reiji Suda; Takayuki Aoki; Shoichi Hirasawa; Akira Nukada; Hiroki Honda; Satoshi Matsuoka

doi:10.1109/ASPDAC.2009.4796483

Aspects of GPU for general purpose high performance computing

Reiji Suda, Takayuki Aoki, Shoichi Hirasawa, Akira Nukada, Hiroki Honda, Satoshi Matsuoka

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

18 Scopus citations

Abstract

We discuss hardware and software aspects of GPGPU, specifically focusing on NVIDIA cards and CUDA, from the viewpoints of parallel computing. The major weak points of GPU against newest supercomputers are identified to be and summarized as only four points: large SIMD vector length, small memory, absence of fast L2 cache, and high register spill penalty. As software concerns, we derive optimal scheduling algorithm for latency hiding of host-device data transfer, and discuss SPMD parallelism on GPUs.

Original language	English
Title of host publication	Proceedings of the ASP-DAC 2009
Subtitle of host publication	Asia and South Pacific Design Automation Conference 2009
Pages	216-223
Number of pages	8
DOIs	https://doi.org/10.1109/ASPDAC.2009.4796483
State	Published - 2009
Externally published	Yes
Event	Asia and South Pacific Design Automation Conference 2009, ASP-DAC 2009 - Yokohama, Japan Duration: 19 Jan 2009 → 22 Jan 2009

Publication series

Name	Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC

Conference

Conference	Asia and South Pacific Design Automation Conference 2009, ASP-DAC 2009
Country/Territory	Japan
City	Yokohama
Period	19/01/09 → 22/01/09

Access to Document

10.1109/ASPDAC.2009.4796483

Cite this

Suda, R., Aoki, T., Hirasawa, S., Nukada, A., Honda, H., & Matsuoka, S. (2009). Aspects of GPU for general purpose high performance computing. In Proceedings of the ASP-DAC 2009: Asia and South Pacific Design Automation Conference 2009 (pp. 216-223). Article 4796483 (Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC). https://doi.org/10.1109/ASPDAC.2009.4796483

@inproceedings{7d572a8b479f4086ab778834c7a6e6e3,

title = "Aspects of GPU for general purpose high performance computing",

abstract = "We discuss hardware and software aspects of GPGPU, specifically focusing on NVIDIA cards and CUDA, from the viewpoints of parallel computing. The major weak points of GPU against newest supercomputers are identified to be and summarized as only four points: large SIMD vector length, small memory, absence of fast L2 cache, and high register spill penalty. As software concerns, we derive optimal scheduling algorithm for latency hiding of host-device data transfer, and discuss SPMD parallelism on GPUs.",

author = "Reiji Suda and Takayuki Aoki and Shoichi Hirasawa and Akira Nukada and Hiroki Honda and Satoshi Matsuoka",

year = "2009",

doi = "10.1109/ASPDAC.2009.4796483",

language = "English",

isbn = "9781424427482",

series = "Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC",

pages = "216--223",

booktitle = "Proceedings of the ASP-DAC 2009",

note = "Asia and South Pacific Design Automation Conference 2009, ASP-DAC 2009 ; Conference date: 19-01-2009 Through 22-01-2009",

}

Suda, R, Aoki, T, Hirasawa, S, Nukada, A, Honda, H & Matsuoka, S 2009, Aspects of GPU for general purpose high performance computing. in Proceedings of the ASP-DAC 2009: Asia and South Pacific Design Automation Conference 2009., 4796483, Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC, pp. 216-223, Asia and South Pacific Design Automation Conference 2009, ASP-DAC 2009, Yokohama, Japan, 19/01/09. https://doi.org/10.1109/ASPDAC.2009.4796483

Aspects of GPU for general purpose high performance computing. / Suda, Reiji; Aoki, Takayuki; Hirasawa, Shoichi et al.
Proceedings of the ASP-DAC 2009: Asia and South Pacific Design Automation Conference 2009. 2009. p. 216-223 4796483 (Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC).

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

TY - GEN

T1 - Aspects of GPU for general purpose high performance computing

AU - Suda, Reiji

AU - Aoki, Takayuki

AU - Hirasawa, Shoichi

AU - Nukada, Akira

AU - Honda, Hiroki

AU - Matsuoka, Satoshi

PY - 2009

Y1 - 2009

N2 - We discuss hardware and software aspects of GPGPU, specifically focusing on NVIDIA cards and CUDA, from the viewpoints of parallel computing. The major weak points of GPU against newest supercomputers are identified to be and summarized as only four points: large SIMD vector length, small memory, absence of fast L2 cache, and high register spill penalty. As software concerns, we derive optimal scheduling algorithm for latency hiding of host-device data transfer, and discuss SPMD parallelism on GPUs.

AB - We discuss hardware and software aspects of GPGPU, specifically focusing on NVIDIA cards and CUDA, from the viewpoints of parallel computing. The major weak points of GPU against newest supercomputers are identified to be and summarized as only four points: large SIMD vector length, small memory, absence of fast L2 cache, and high register spill penalty. As software concerns, we derive optimal scheduling algorithm for latency hiding of host-device data transfer, and discuss SPMD parallelism on GPUs.

UR - http://www.scopus.com/inward/record.url?scp=64549153553&partnerID=8YFLogxK

U2 - 10.1109/ASPDAC.2009.4796483

DO - 10.1109/ASPDAC.2009.4796483

M3 - Conference contribution

AN - SCOPUS:64549153553

SN - 9781424427482

T3 - Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC

SP - 216

EP - 223

BT - Proceedings of the ASP-DAC 2009

T2 - Asia and South Pacific Design Automation Conference 2009, ASP-DAC 2009

Y2 - 19 January 2009 through 22 January 2009

ER -

Aspects of GPU for general purpose high performance computing

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this