Modeling optimization of stencil computations via domain-level properties

Modeling optimization of stencil computations via domain-level properties
复制标题

通过域级属性对模板计算进行建模优化

DOI:
10.1145/3528425.3529103
复制
发表时间:
2022
期刊:
Thirteenth International Workshop on Programming Models and Applications for Multicores and Manycores
影响因子:
--
通讯作者:
Runnels, Brandon
Runnels, Brandon
中科院分区:
--
文献类型:
--
作者:
Nesterenko, Brandon;Yi, Qing;Lin, Pei-Hung;Liao, Chunhua;Runnels, Brandon

文献摘要

参考文献

相似文献

模板计算在科学模拟领域有着广泛的应用,其性能对许多大规模数值应用的整体效率至关重要。存在许多优化技术,其中大多数是不同的平铺和并行化策略,以系统地提高模板计算的效率。然而,这些优化的有效性明显不同,这取决于不同模板所表现出的广泛的性质。本文研究了几种著名的模板优化策略,并提出了一种新的方法来有效地指导这些优化的组合,通过建模这些优化与模板的四个领域级属性:空间维度、时间顺序、精度顺序和方向相关性。当使用我们的预测模型来指导对五个真实模板应用程序的优化时,我们能够识别出性能比两个高度优化的模板库平均高出2.4倍的优化策略。
Stencil computations are widely used in the scientific simulation domain, and their performance is critical to the overall efficiency of many large-scale numerical applications. Many optimization techniques, most of them varying strategies of tiling and parallelization, exist to systematically enhance the efficiency of stencil computations. However, the effectiveness of these optimizations vary significantly depending on the wide range of properties demonstrated by the different stencils. This paper studies several well-known optimization strategies for stencils and presents a new approach to effectively guide the composition of these optimizations, by modeling their interactions with four domain-level properties of stencils: spatial dimensionality, temporal order, order of accuracy, and directional dependence. When using our prediction model to guide optimizations for five real-world stencil applications, we were able to identify optimization strategies that outperformed two highly optimized stencil libraries by an average of 2.4x.
Snowflake:轻量级便携式模板 DSL
DOI: --
发表时间: 2017
期刊: IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum
影响因子: --
作者:
Nathan Zhang;Michael B. Driscoll;Chick Markley;Samuel Williams;P. Basu;A. Fox
通讯作者: A. Fox
使用 Kerncraft 自动循环内核分析和性能建模
DOI: --
发表时间: 2015
期刊: International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems
影响因子: --
作者:
Julian Hammer;G. Hager;Jan Eitzinger;G. Wellein
通讯作者: G. Wellein
电磁学第 1 卷(测试版)
DOI: --
发表时间: 2018
期刊:
影响因子: --
作者:
S. Ellingson
通讯作者: S. Ellingson
DOI: 10.1109/pmbs.2018.8641578
发表时间: 2018
期刊: 2018 IEEE/ACM Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS)
影响因子: --
作者:
Jan Laukemann;Julian Hammer;Johannes Hofmann;G. Hager;G. Wellein
通讯作者: G. Wellein
POET:经验调优的参数化优化
DOI: 10.1109/ipdps.2007.370637
发表时间: 2007
期刊: 2007 IEEE International Parallel and Distributed Processing Symposium
影响因子: --
作者:
Qing Yi;Keith Seymour;Haihang You;R. Vuduc;D. Quinlan
通讯作者: D. Quinlan