Power and performance characterization and modeling of GPU-accelerated systems

Yuki Abe, Hiroshi Sasaki, Shinpei Kato, Koji Inoue, Masato Edahiro, Martin Peres

研究成果: Conference contribution

48 被引用数 (Scopus)

抄録

Graphics processing units (GPUs) provide an order-of-magnitude improvement on peak performance and performance-per-watt as compared to traditional multicore CPUs. However, GPU-accelerated systems currently lack a generalized method of power and performance prediction, which prevents system designers from an ultimate goal of dynamic power and performance optimization. This is due to the fact that their power and performance characteristics are not well captured across architectures, and as a result, existing power and performance modeling approaches are only available for a limited range of particular GPUs. In this paper, we present power and performance characterization and modeling of GPU-accelerated systems across multiple generations of architectures. Characterization and modeling both play a vital role in optimization and prediction of GPU-accelerated systems. We quantify the impact of voltage and frequency scaling on each architecture with a particularly intriguing result that a cutting-edge Kepler-based GPU achieves energy saving of 75% by lowering GPU clocks in the best scenario, while Fermi- and Tesla-based GPUs achieve no greater than 40% and 13%, respectively. Considering these characteristics, we provide statistical power and performance modeling of GPU-accelerated systems simplified enough to be applicable for multiple generations of architectures. One of our findings is that even simplified statistical models are able to predict power and performance of cutting-edge GPUs within errors of 20% to 30% for any set of voltage and frequency pair.

本文言語English
ホスト出版物のタイトルProceedings - IEEE 28th International Parallel and Distributed Processing Symposium, IPDPS 2014
出版社IEEE Computer Society
ページ113-122
ページ数10
ISBN(印刷版)9780769552071
DOI
出版ステータスPublished - 2014
外部発表はい
イベント28th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2014 - Phoenix, AZ, United States
継続期間: 2014 5月 192014 5月 23

出版物シリーズ

名前Proceedings of the International Parallel and Distributed Processing Symposium, IPDPS
ISSN(印刷版)1530-2075
ISSN(電子版)2332-1237

Conference

Conference28th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2014
国/地域United States
CityPhoenix, AZ
Period14/5/1914/5/23

ASJC Scopus subject areas

  • 計算理論と計算数学
  • コンピュータ ネットワークおよび通信
  • ハードウェアとアーキテクチャ
  • ソフトウェア

フィンガープリント

「Power and performance characterization and modeling of GPU-accelerated systems」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル