Asynchronous Hybrid Parallel SpMV in an Industrial Application

被引:0
|
作者
Thiele, Christopher [1 ]
Araya-Polo, Mauricio [2 ]
Stoyanov, Dimitar [1 ]
Frank, Florian [3 ]
Alpak, Faruk O. [2 ]
机构
[1] Fraunhofer ITWM, Fraunhofer Pl 1, D-67663 Kaiserslautern, Germany
[2] Shell Int Explorat & Prod Inc, 3333 Highway 6 South, Houston, TX 77082 USA
[3] Rice Univ, 6100 Main St, Houston, TX 77005 USA
关键词
Linear systems; Sparse matrices; Fluid dynamics; Numerical simulation; Hydrocarbons;
D O I
10.1109/CSCI.2016.225
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Pore-scale simulation is both computationally challenging and a key element of hydrocarbon exploration and production ([1], [2], [3]). The subject of this study is the efficient solution of the linear systems arising from the discretization of the Cahn- Hilliard equation ([4], [5]), which governs the separation of a two- component fluid mixture. We investigate synchronous and asynchronous parallel iterative solvers for our simulation. We focus our efforts on two main aspects: first, software infrastructure that supports this kind of computations, mainly Trilinos and GaspiLS. Second, scalability and performance analysis. The asynchronous solver performs well and scales better than the synchronous ones for most test problems, especially for large systems distributed over many computational nodes (up to 23% faster and 10% higher parallel efficiency in these cases).
引用
收藏
页码:1196 / 1201
页数:6
相关论文
共 50 条
  • [21] An effective SPMV based on block strategy and hybrid compression on GPU
    Huanyu Cui
    Nianbin Wang
    Yuhua Wang
    Qilong Han
    Yuezhu Xu
    The Journal of Supercomputing, 2022, 78 : 6318 - 6339
  • [22] Hybrid Parallel Stochastic Configuration Networks for Industrial Data Analytics
    Dai, Wei
    Zhou, Xinyu
    Li, Depeng
    Zhu, Song
    Wang, Xuesong
    IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2022, 18 (04) : 2331 - 2341
  • [23] An effective SPMV based on block strategy and hybrid compression on GPU
    Cui, Huanyu
    Wang, Nianbin
    Wang, Yuhua
    Han, Qilong
    Xu, Yuezhu
    JOURNAL OF SUPERCOMPUTING, 2022, 78 (05): : 6318 - 6339
  • [24] Scalability of Hybrid Sparse Matrix Dense Vector (SpMV) Multiplication
    Page, Brian A.
    Kogge, Peter M.
    PROCEEDINGS 2018 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS), 2018, : 406 - 414
  • [25] Innovative hybrid welding process in an industrial application
    Kaierle, S
    Bongard, K
    Dahmen, M
    Poprawe, R
    ICALEO(R) 2000: PROCEEDINGS OF THE LASER MATERIALS PROCESSING CONFERENCE, VOL 89, 2000, 89 : C91 - C98
  • [26] Hybrid asynchronous algorithm for parallel kinetic Monte Carlo simulations of thin film growth
    Shim, Y
    Amar, JG
    JOURNAL OF COMPUTATIONAL PHYSICS, 2006, 212 (01) : 305 - 317
  • [27] Impacts of Three Soft-Fault Models on Hybrid Parallel Asynchronous Iterative Methods
    Coleman, Evan
    Jensen, Erik J.
    Sosonkina, Masha
    2018 30TH INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD 2018), 2018, : 458 - 465
  • [28] Optimization of ply angles in laminated composite structures by a hybrid, asynchronous, parallel evolutionary algorithm
    Keller, David
    COMPOSITE STRUCTURES, 2010, 92 (11) : 2781 - 2790
  • [29] A parallel computing method using blocked format with optimal partitioning for SpMV on GPU
    Yang, Wangdong
    Li, Kenli
    Li, Keqin
    JOURNAL OF COMPUTER AND SYSTEM SCIENCES, 2018, 92 : 152 - 170
  • [30] Parallel asynchronous algorithms: A survey
    Spiteri, Pierre
    ADVANCES IN ENGINEERING SOFTWARE, 2020, 149