High level implementation of geometric multigrid solvers for finite element problems: Applications in atmospheric modelling

High level implementation of geometric multigrid solvers for finite element problems: Applications in atmospheric modelling
复制标题

有限元问题几何多重网格求解器的高级实现:在大气建模中的应用

DOI:
10.1016/j.jcp.2016.09.037
复制
发表时间:
2016
影响因子:
4.1
通讯作者:
Mitchell L
Mitchell L
中科院分区:
物理与天体物理2区
文献类型:
--
作者:
Mitchell L

文献摘要

参考文献

被引文献

相似文献

椭圆型偏微分方程(PDE)的多重网格预处理子的实现是一个挑战,因为所得到的算法和相应的计算机代码是复杂的。对于非结构化网格上复杂的(混合)有限元离散,高效的实现可能非常耗时,并且要求程序员在多核CPU上具有深入的数学理论、并行计算和优化技术的知识。在这篇文章中,我们展示了如何通过使用一个框架来显着简化定制的多重网格预处理器的开发,该框架允许在正确的抽象级别上表达算法的每个组成部分。我们的方法(1)允许用接近问题的数学公式的语言来表示有限元问题,(2)保证自动生成和有效执行并行优化的低级计算机代码,(3)足够灵活地支持不同的抽象级别,并使程序员能够控制预条件器的细节。我们使用Firedrake/PyOP2软件包的可组合抽象来演示该方法在大气模拟中求解强各向异性偏微分方程组的有效性。PDE的弱形式用统一形式语言(UFL)表示,较低的PyOP2抽象层允许为定制的几何多重网格预处理器手动设计计算内核。我们将该预处理器的性能与单层方法和HYPRE的Bomer AMG算法进行了比较。Firedrake/PyOP2代码本质上是并行的,我们对Archer超级计算机上的单个节点(24核)进行了详细的性能分析。我们的实施使用了很大一部分可用内存带宽,并且在高达6,144个计算核心上显示了非常好的弱伸缩性。
The implementation of efficient multigrid preconditioners for elliptic partial differential equations (PDEs) is a challenge due to the complexity of the resulting algorithms and corresponding computer code. For sophisticated (mixed) finite element discretisations on unstructured grids an efficient implementation can be very time consuming and requires the programmer to have in-depth knowledge of the mathematical theory, parallel computing and optimisation techniques on manycore CPUs. In this paper we show how the development of bespoke multigrid preconditioners can be simplified significantly by using a framework which allows the expression of the each component of the algorithm at the correct abstraction level. Our approach (1) allows the expression of the finite element problem in a language which is close to the mathematical formulation of the problem, (2) guarantees the automatic generation and efficient execution of parallel optimised low-level computer code and (3) is flexible enough to support different abstraction levels and give the programmer control over details of the preconditioner. We use the composable abstractions of the Firedrake/PyOP2 package to demonstrate the efficiency of this approach for the solution of strongly anisotropic PDEs in atmospheric modelling. The weak formulation of the PDE is expressed in Unified Form Language (UFL) and the lower PyOP2 abstraction layer allows the manual design of computational kernels for a bespoke geometric multigrid preconditioner. We compare the performance of this preconditioner to a single-level method and hypre's BoomerAMG algorithm. The Firedrake/PyOP2 code is inherently parallel and we present a detailed performance analysis for a single node (24 cores) on the ARCHER supercomputer. Our implementation utilises a significant fraction of the available memory bandwidth and shows very good weak scaling on up to 6,144 compute cores.
ufl:统一形式语言
DOI: --
发表时间: 2016
期刊:
影响因子: --
作者:
M. Alnæs;Miklós Homolya;G. N. Wells;A. Logg;Johannes Ring;M. Rognes;chrisrichardson;D. Ham;Aslak W. Bergersen;K. Mardal;L. Mitchell
通讯作者: L. Mitchell
COFFEE:快速表达式评估的编译器
DOI: 10.5281/zenodo.56636
发表时间: 2016
期刊: ACM Transactions on Mathematical Software (TOMS)
影响因子: --
作者:
F. Luporini;Michael Lange;F. P. Russell;G. Markall;Miklós Homolya;L. Mitchell;D. Ham;Florian Rathgeber
通讯作者: Florian Rathgeber
fiat:有限元自动制表器
DOI: 10.5281/zenodo.46580
发表时间: 2016
影响因子: 3.2
作者:
M. Rognes;C. Cotter;R. Kirby;G. N. Wells;A. Logg;Lizao Li;Johannes Ring;Jan Blechta;Miklós Homolya;M. Alnæs;N. Schlömer;D. Ham;mliertzer;Aslak W. Bergersen;Andrew T. T. McRae;Florian Rathgeber;L. Mitchell
通讯作者: L. Mitchell
DOI: 10.1145/2687415
发表时间: 2014
期刊: ACM Transactions on Architecture and Code Optimization (TACO)
影响因子: --
作者:
F. Luporini;A. Varbanescu;Florian Rathgeber;Gheorghe;J. Ramanujam;D. Ham;P. Kelly
通讯作者: P. Kelly
DOI: 10.1007/s00607-010-0110-3
发表时间: 2010-11
期刊: Computing
影响因子: 3.7
作者:
A. Dedner;R. Klöfkorn;M. Nolte;Mario Ohlberger
通讯作者: A. Dedner;R. Klöfkorn;M. Nolte;Mario Ohlberger