Automatic Hybrid OpenMP plus MPI Program Generation for Dynamic Programming Problems

被引:0
|
作者
VandenBerg, Denny R. [1 ]
Stout, Quentin F. [1 ]
机构
[1] Univ Michigan, Ann Arbor, MI 48105 USA
关键词
dynamic programming; automatic program generation; hybrid; openmp; mpi; bandit problem;
D O I
10.1109/CLUSTER.2011.28
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
We describe a program that automatically generates a hybrid OpenMP + MPI program for a class of recursive calculations with template dependencies. Many useful generalized dynamic programming problems fit this category, such as Multiple String Alignment and multi-arm Bernoulli Bandit problems. Solving problems like these, especially those involving several dimensions, can use a significant amount of memory and time. Our generator addresses these issues by dividing the problem into many tiles that can be solved in parallel. Programs generated using this program generator are capable of solving large problems and achieve good scalability when run on many cores. The input supplied to the generator is a high level description of the problem; the output is a fully functioning parallel program for a cluster of shared memory nodes. This high level approach to parallel computation allows the generator to have a large amount of control over memory allocation, load balancing and calculation ordering.
引用
收藏
页码:178 / 186
页数:9
相关论文
共 50 条
  • [1] Automatic Hybrid MPI plus OpenMP Code Generation with 11c
    Reyes, Ruyman
    Dorta, Antonio J.
    Almeida, Francisco
    de Sande, Francisco
    [J]. RECENT ADVANCES IN PARALLEL VIRTUAL MACHINE AND MESSAGE PASSING INTERFACE, PROCEEDINGS, 2009, 5759 : 185 - 195
  • [2] Parallel programming for OSEM reconstruction with MPI, OpenMP, and hybrid MPI-OpenMP
    Jones, MD
    Yao, RT
    [J]. 2004 IEEE NUCLEAR SCIENCE SYMPOSIUM CONFERENCE RECORD, VOLS 1-7, 2004, : 3036 - 3042
  • [3] Automatic Partitioning of MPI Operations in MPI plus OpenMP Applications
    Jammer, Tim
    Bischof, Christian
    [J]. HIGH PERFORMANCE COMPUTING - ISC HIGH PERFORMANCE DIGITAL 2021 INTERNATIONAL WORKSHOPS, 2021, 12761 : 191 - 198
  • [4] Early experiments with the OpenMP/MPI hybrid programming model
    Lusk, Ewing
    Chan, Anthony
    [J]. OPENMP IN A NEW ERA OF PARALLELISM, PROCEEDINGS, 2008, 5004 : 36 - +
  • [5] Parallelization of Array Method with Hybrid Programming: OpenMP and MPI
    Velarde Martinez, Apolinar
    [J]. APPLIED SCIENCES-BASEL, 2022, 12 (15):
  • [6] Hybrid MPI/OpenMP Programming on the Tilera Manycore Architecture
    Chandru, Vishwanathan
    Mueller, Frank
    [J]. 2016 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS 2016), 2016, : 326 - 333
  • [7] Toward Heterogeneous MPI plus MPI Programming: Comparison of OpenMP and MPI Shared Memory Models
    Szustak, Lukasz
    Wyrzykowski, Roman
    Halbiniak, Kamil
    Bratek, Pawel
    [J]. EURO-PAR 2019: PARALLEL PROCESSING WORKSHOPS, 2020, 11997 : 270 - 281
  • [8] A Framework for an Automatic Hybrid MPI plus OpenlVIP code generation
    Hamidouche, Khaled
    Falcou, Joel
    Etiemble, Daniel
    [J]. HIGH PERFORMANCE COMPUTING SYMPOSIUM 2011 (HPC 2011) - 2011 SPRING SIMULATION MULTICONFERENCE - BK 6 OF 8, 2011, 43 (02): : 48 - 55
  • [9] Automatic performance analysis of hybrid MPI/OpenMP applications
    Wolf, F
    Mohr, B
    [J]. ELEVENTH EUROMICRO CONFERENCE ON PARALLEL, DISTRIBUTED AND NETWORK-BASED PROCESSING, PROCEEDINGS, 2003, : 13 - 22
  • [10] Automatic performance analysis of hybrid MPI/OpenMP applications
    Wolf, F
    Mohr, B
    [J]. JOURNAL OF SYSTEMS ARCHITECTURE, 2003, 49 (10-11) : 421 - 439