Scaling Study of Flow Simulations on Composable Cyberinfrastructure

Scaling Study of Flow Simulations on Composable Cyberinfrastructure
复制标题

可组合网络基础设施流模拟的规模化研究

DOI:
10.1145/3569951.3597565
复制
发表时间:
2023
期刊:
USA
影响因子:
--
通讯作者:
Dang, Francis
Dang, Francis
中科院分区:
--
文献类型:
--
作者:
Mishra, Sambit;Witherden, Freddie;Chakravorty, Dhruva;Perez, Lisa;Dang, Francis

文献摘要

参考文献

被引文献

相似文献

采用可组合方法的网络基础设施 (CI) 系统使研究人员能够定义最适合满足其计算工作流程需求的资源。在这些方法中,通过软件定义的网络组合分散的计算资源有望支持需要动态访问大量加速器或内存的工作负载。关于这种方法的架构引起的限制,以及它如何影响科学和工程应用软件,还有很多需要理解的地方。在这里,我们研究了高度可扩展的开源流求解器 PyFR 在使用软件定义的 PCIe Gen4 结构编排的基于 GPU 的可组合环境中的性能。 PyFR 强调 GPU 之间的通信,并有助于了解如何优化配置分类资源上的 GPU 以获得性能。将组合配置的强扩展和弱扩展性能研究与具有 InfiniBand 互连的传统 CPU-GPU 集群进行比较。讨论了影响性能的因素以及可组合设备对新基准套件的需求。
Cyberinfrastructure (CI) systems employing composable approaches give researchers the capability to define resources best suited to meet the needs of their computational workflows. Among these approaches, composing disaggregated computing resources over a software-defined network offers the promise of supporting workloads requiring dynamic access to a large pool of accelerators or memory. Much remains to be understood about the architecture-induced constraints of this approach, and how it impacts scientific and engineering applications software. Here, we study the performance of the highly scalable open-source flow solver, PyFR, in a GPU-based composable environment orchestrated using a software-defined PCIe Gen4 fabric. PyFR emphasizes communication between GPUs and helps understand how GPUs on disaggregated resources can be optimally configured for performance. Strong-scaling and weak-scaling performance studies on composed configurations are compared to a traditional CPU-GPU cluster with InfiniBand interconnect. Factors affecting performance, and the need for new benchmark suites for composable devices are discussed.
开发可组合网络基础设施的综合应用基准:GPU 上扩展分子动力学应用的研究
DOI: 10.1145/3569951.3597556
发表时间: 2023
期刊: USA
影响因子: --
作者:
Lawrence, Richard E;Chakravorty, Dhruva K;He, Zhenhua;Dang, Francis;Perez, Lisa M;Brashear, Wesley A;Liu, Honggao
通讯作者: Liu, Honggao
泰勒-格林涡作为使用低马赫求解器进行高保真燃烧模拟的基准
DOI: 10.1016/j.compfluid.2021.104935
发表时间: 2021
期刊: Computers & Fluids
影响因子: 2.8
作者:
A. Abdelsamie;G. Lartigue;C. Frouzakis;D. Thévenin
通讯作者: D. Thévenin
DOI: 10.1177/10943420211020803
发表时间: 2021-06
期刊: The International Journal of High Performance Computing Applications
影响因子: --
作者:
T. Kolev;P. Fischer;M. Min;J. Dongarra;Jed Brown;V. Dobrev;T. Warburton;S. Tomov;M. Shephard;A. Abdelfattah;V. Barra;Natalie N. Beams;Jean-Sylvain Camier;N. Chalmers;Yohann Dudouit;A. Karakus;I. Karlin;S. Kerkemeier;Yu-Hsiang Lan;David S. Medina;E. Merzari;A. Obabko;Will Pazner;T. Rathnayake;Cameron W. Smith;Lukas Spies;K. Swirydowicz;Jeremy L. Thompson;A. Tomboulides;V. Tomov
通讯作者: T. Kolev;P. Fischer;M. Min;J. Dongarra;Jed Brown;V. Dobrev;T. Warburton;S. Tomov;M. Shephard;A. Abdelfattah;V. Barra;Natalie N. Beams;Jean-Sylvain Camier;N. Chalmers;Yohann Dudouit;A. Karakus;I. Karlin;S. Kerkemeier;Yu-Hsiang Lan;David S. Medina;E. Merzari;A. Obabko;Will Pazner;T. Rathnayake;Cameron W. Smith;Lukas Spies;K. Swirydowicz;Jeremy L. Thompson;A. Tomboulides;V. Tomov
可组合网络基础设施上分布式深度学习工作负载的性能
DOI: 10.1145/3569951.3593601
发表时间: 2023
期刊: USA.
影响因子: --
作者:
He, Zhenhua;Saluja, Aditi;Lawrence, Richard;Chakravorty, Dhruva;Dang, Francis;Perez, Lisa;Liu, Honggao
通讯作者: Liu, Honggao