Performance and Energy Analysis of OpenMP Runtime Systems with Dense Linear Algebra Algorithms

被引：7
|
作者：
Lima, Joao V. F. [1 ]
Rais, Issam [2 ]
Lefevre, Laurent [2 ]
Gautier, Thierry [2 ]
机构：
[1] Univ Fed Santa Maria, Santa Maria, RS, Brazil
[2] Univ Claude Bernard Lyon 1, Univ Lyon, CNRS, INRIA,ENS Lyon,LIP, Villeurbanne, France
来源：
2017 INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING WORKSHOPS (SBAC-PADW) | 2017年
关键词：
D O I：
10.1109/SBAC-PADW.2017.10
中图分类号：
TP3 [计算技术、计算机技术];
学科分类号：
0812 ;
摘要：
In this paper, we analyse performance and energy consumption of four OpenMP runtime systems over a NUMA platform. We present an experimental study to characterize OpenMP runtime systems on the three main kernels in dense linear algebra algorithms (Cholesky, LU and QR) in terms of performance and energy consumption. Our experimental results suggest that OpenMP runtime systems can be considered as a new energy leverage. For instance, a LU factorization with concurrent write extension from libKOMP achieved up to 1.75 of performance gain and 1.56 of energy decrease.
引用
收藏
页码：7 / 12
页数：6
相关论文
共 50 条
  • [1] Performance and energy analysis of OpenMP runtime systems with dense linear algebra algorithms
    Ferreira Lima, Joao Vicente
    Rais, Issam
    Lefevre, Laurent
    Gautier, Thierry
    INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS, 2019, 33 (03): : 431 - 443
  • [2] Enhancing performance and energy consumption of runtime schedulers for dense linear algebra
    Alonso, Pedro
    Dolz, Manuel F.
    Igual, Francisco D.
    Mayo, Rafael
    Quintana-Orti, Enrique S.
    CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2014, 26 (15): : 2591 - 2611
  • [3] Performance and energy impact of OpenMP runtime configurations on power constrained systems
    Bari, Md Abdullah Shahneous
    Malik, Abid M.
    Qawasmeh, Ahmad
    Chapman, Barbara
    SUSTAINABLE COMPUTING-INFORMATICS & SYSTEMS, 2019, 23 : 1 - 12
  • [4] Profiling high performance dense linear algebra algorithms on multicore architectures for power and energy efficiency
    Ltaief, Hatem
    Luszczek, Piotr
    Dongarra, Jack
    COMPUTER SCIENCE-RESEARCH AND DEVELOPMENT, 2012, 27 (04): : 277 - 287
  • [5] FLOPs as a Discriminant for Dense Linear Algebra Algorithms
    Lopez, Francisco
    Karlsson, Lars
    Bientinesi, Paolo
    51ST INTERNATIONAL CONFERENCE ON PARALLEL PROCESSING, ICPP 2022, 2022,
  • [6] PARALLEL ALGORITHMS FOR DENSE LINEAR ALGEBRA COMPUTATIONS
    GALLIVAN, KA
    PLEMMONS, RJ
    SAMEH, AH
    SIAM REVIEW, 1990, 32 (01) : 54 - 135
  • [7] The science of deriving dense linear algebra algorithms
    Bientinesi, P
    Gunnels, JA
    Myers, ME
    Quintana-Orti, ES
    Van De Geijn, RA
    ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE, 2005, 31 (01): : 1 - 26
  • [8] IMPROVING PERFORMANCE OF LINEAR ALGEBRA ALGORITHMS FOR DENSE MATRICES, USING ALGORITHMIC PREFETCH
    AGARWAL, RC
    GUSTAVSON, FG
    ZUBAIR, M
    IBM JOURNAL OF RESEARCH AND DEVELOPMENT, 1994, 38 (03) : 265 - 275
  • [9] Performance Modeling for Dense Linear Algebra
    Peise, Elmar
    Bientinesi, Paolo
    2012 SC COMPANION: HIGH PERFORMANCE COMPUTING, NETWORKING, STORAGE AND ANALYSIS (SCC), 2012, : 406 - 416
  • [10] Analysis of dynamically scheduled tile algorithms for dense linear algebra on multicore architectures
    Haidar, Azzam
    Ltaief, Hatem
    YarKhan, Asim
    Dongarra, Jack
    CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2012, 24 (03): : 305 - 321
12345