Details

Autor(en) / Beteiligte

• Lai, Bo-Cheng
• Wu, Tung-Yu
• Chiu, Tsou-Han
• Li, Kun-Chun
• Lee, Chia-Ying
• Chien, Wei-Chen
• Wong, Wing Hung

Titel

Towards high performance data analytic on heterogeneous many-core systems: A study on Bayesian Sequential Partitioning

Ist Teil von

• Journal of parallel and distributed computing, 2018-12, Vol.122, p.36-50

Ort / Verlag

United States: Elsevier Inc

Erscheinungsjahr

2018

Link zum Volltext

Quelle

ScienceDirect Journals (5 years ago - present)

Beschreibungen/Notizen

• Bayesian Sequential Partitioning (BSP) is a statistically effective density estimation method to comprehend the characteristics of a high dimensional data space. The intensive computation of the statistical model and the counting of enormous data have caused serious design challenges for BSP to handle the growing volume of the data. This paper proposes a high performance design of BSP by leveraging a heterogeneous CPU/GPGPU system that consists of a host CPU and a K80 GPGPU. A series of techniques, on both data structures and execution management policies, is implemented to extensively exploit the computation capability of the heterogeneous many-core system and alleviate system bottlenecks. When compared with a parallel design on a high-end CPU, the proposed techniques achieve 48x average runtime enhancement while the maximum speedup can reach 78.76x. •Techniques to speedup Bayesian Sequential Partitioning by 48x on a heterogeneous many-core system.•Proposes a series of techniques, for both data structures and execution management policies.•Achieve 106x average runtime enhancement while the maximum speedup can reach 197.96x.