EAGER: Assessment of the Numerical Reproducibility in Large-Scale Scientific Simulations on Multicore Architectures
EAGER: Assessment of the Numerical Reproducibility in Large-Scale Scientific Simulations on Multicore Architectures
批准号:
1446794
负责人:
Michela Taufer
金额:
$9.0万
依托单位:
依托单位国家:
美国
项目类别:
Standard Grant
财政年份:
2014
资助国家:
美国
项目状态:
已结题
起止时间:
2014-06-15 至 2016-05-31
中文摘要
执行并发性的趋势为开发能够自动有效地建模和减轻千万亿次架构之外的不可再现性并进入亿亿次的方法提供了一个令人信服的案例。预计艾级的高性能计算机将表现出巨大的并发水平-比当前平台大10,000倍-这将使计算机模拟从批量同步执行转向多线程方法和异步I/O。模拟计算和分析例程也将在亿级平台上紧密耦合,这需要这两个工作流组件以极高的并发水平工作。随着并发级别的增加,舍入误差对数值再现性的影响也会增加,最终影响科学模拟再现程序执行和数值结果的能力。在这种情况下,不可重现的结果可能不被期望可重现行为的科学界所信任,任何追求可重现性的尝试都可能以过高的性能为代价。这个“高风险-高回报”项目研究了在尖端多核架构中,当并发执行突发和工作流确定性消失时,舍入误差对结果可重现性的影响。为此,该项目采用一种称为“复合精度浮点运算”的数学方法对科学应用中的舍入误差进行建模,并展示了这种方法如何减轻误差漂移。 在前期工作中使用的基准测试套件进行了扩展,以涵盖更大范围的应用程序的模式,并用于评估新一代多核架构的复合精度的缓解影响。最后,该项目量化了所提出的方法的成本和缓解因素,以缓解不同基准和平台的误差传播。该项目将通过开发和推广有效的软件解决方案,为广泛的应用程序及其代码在大规模并行环境下以高度并发执行时的舍入误差传播问题,提高对艾级数值再现性的认识和理解系统.
英文摘要
Trends in execution concurrency make a compelling case for the development of methods able to automatically and efficiently model and mitigate irreproducibility beyond petascale architectures and into the exascale. It is expected that high performance computers at the exascale will exhibit a massively large level of concurrency - a factor of 10,000 greater than on current platforms - which will move computer simulations from bulk-synchronous executions to multithreading approaches and asynchronous I/O. Simulation calculations and analysis routines will also be tightly coupled on exascale platforms, requiring these two workflow components to work at extremely high levels of concurrency. As concurrency levels increase, the impact of rounding errors on numerical reproducibility also increases, ultimately affecting the ability of scientific simulations to reproduce program executions and numerical results. Under these circumstances, irreproducible results may not be trusted by a scientific community expecting reproducible behaviors and any attempt to pursue reproducibility may come at a cost in performance that is too high.This "high risk-high payoff" project studies the impact of rounding errors on result reproducibility when concurrent executions burst and workflow determinism vanishes in cutting-edge multicore architectures. To this end, the project models rounding-errors in scientific applications with a mathematical method called "composite precision floating-point arithmetic" and shows how this method can mitigate error drifting. A benchmark suite used in preliminary work is extended to cover a larger range of applications' patterns and used to assess the mitigating impact of the composite precision on new generations of multicore architectures. Lastly, the project quantifies the cost and mitigation factors of the proposed method to mitigate error propagations for the diverse benchmarks and platforms.The project will advance knowledge and understanding in numerical reproducibility at the exascale by developing and disseminating effective software solutions to the rounding error propagation problem for a broad set of applications and their codes when executed with high degrees of concurrency on massively parallel systems.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
EAGER: A Comprehensive Approach for Generating, Sharing, Searching, and Using High-Resolution Terrain Parameters
-
批准号:2334945
-
项目类别:Standard Grant
-
资助金额:$22.5万
-
财政年份:2023
-
负责人:Michela Taufer
-
依托单位:
Collaborative Research: SHF: Small: Model-driven Design and Optimization of Dataflows for Scientific Applications
-
批准号:2331152
-
项目类别:Standard Grant
-
资助金额:$42.4万
-
财政年份:2023
-
负责人:Michela Taufer
-
依托单位:
SHF: Small: Methods, Workflows, and Data Commons for Reducing Training Costs in Neural Architecture Search on High-Performance Computing Platforms
-
批准号:2223704
-
项目类别:Standard Grant
-
资助金额:$62.4万
-
财政年份:2022
-
负责人:Michela Taufer
-
依托单位:
Collaborative Research: Elements: SENSORY: Software Ecosystem for kNowledge diScOveRY - a data-driven framework for soil moisture applications
-
批准号:2103845
-
项目类别:Standard Grant
-
资助金额:$35.0万
-
财政年份:2021
-
负责人:Michela Taufer
-
依托单位:
Collaborative Research: PPoSS: Planning: Performance Scalability, Trust, and Reproducibility: A Community Roadmap to Robust Science in High-throughput Applications
-
批准号:2028923
-
项目类别:Standard Grant
-
资助金额:$9.0万
-
财政年份:2020
-
负责人:Michela Taufer
-
依托单位:
Collaborative Research: EAGER: Advancing Reproducibility in Multi-Messenger Astrophysics
-
批准号:2041977
-
项目类别:Standard Grant
-
资助金额:$10.0万
-
财政年份:2020
-
负责人:Michela Taufer
-
依托单位:
SHF: Medium: Collaborative Research: ANACIN-X: Analysis and modeling of Nondeterminism and Associated Costs in eXtreme scale applications
-
批准号:1900888
-
项目类别:Continuing Grant
-
资助金额:$91.57万
-
财政年份:2019
-
负责人:Michela Taufer
-
依托单位:
Collaborative: EAGER: Exploring and Advancing the State of the Art in Robust Science in Gravitational Wave Physics
-
批准号:1841399
-
项目类别:Standard Grant
-
资助金额:$7.5万
-
财政年份:2018
-
负责人:Michela Taufer
-
依托单位:
Collaborative: EAGER: Exploring and Advancing the State of the Art in Robust Science in Gravitational Wave Physics
-
批准号:1823372
-
项目类别:Standard Grant
-
资助金额:$7.5万
-
财政年份:2018
-
负责人:Michela Taufer
-
依托单位:
SHF:Medium:Collaborative Research:A comprehensive methodology to pursue reproducible accuracy in ensemble scientific simulations on multi- and many-core platforms
-
批准号:1841552
-
项目类别:Standard Grant
-
资助金额:$15.73万
-
财政年份:2018
-
负责人:Michela Taufer
-
依托单位:
BIGDATA: IA: Collaborative Research: In Situ Data Analytics for Next Generation Molecular Dynamics Workflows
-
批准号:1841758
-
项目类别:Standard Grant
-
资助金额:$98.0万
-
财政年份:2018
-
负责人:Michela Taufer
-
依托单位:
CIF21 DIBBs: PD: Cyberinfrastructure Tools for Precision Agriculture in the 21st Century
-
批准号:1854312
-
项目类别:Standard Grant
-
资助金额:$43.81万
-
财政年份:2018
-
负责人:Michela Taufer
-
依托单位:
BIGDATA: IA: Collaborative Research: In Situ Data Analytics for Next Generation Molecular Dynamics Workflows
-
批准号:1741057
-
项目类别:Standard Grant
-
资助金额:$98.0万
-
财政年份:2017
-
负责人:Michela Taufer
-
依托单位:
CIF21 DIBBs: PD: Cyberinfrastructure Tools for Precision Agriculture in the 21st Century
-
批准号:1724843
-
项目类别:Standard Grant
-
资助金额:$50.0万
-
财政年份:2017
-
负责人:Michela Taufer
-
依托单位:
Student Support: IEEE Cluster 2017 Conference
-
批准号:1648617
-
项目类别:Standard Grant
-
资助金额:$2.0万
-
财政年份:2016
-
负责人:Michela Taufer
-
依托单位:
Student Support: IEEE Cluster 2015-2016 Conferences
-
批准号:1550348
-
项目类别:Standard Grant
-
资助金额:$2.0万
-
财政年份:2015
-
负责人:Michela Taufer
-
依托单位:
SHF:Medium:Collaborative Research:A comprehensive methodology to pursue reproducible accuracy in ensemble scientific simulations on multi- and many-core platforms
-
批准号:1513025
-
项目类别:Standard Grant
-
资助金额:$42.79万
-
财政年份:2015
-
负责人:Michela Taufer
-
依托单位:
Student Support: IEEE Cluster 2014 Conference; Madrid Spain; September 22-26, 2014
-
批准号:1441397
-
项目类别:Standard Grant
-
资助金额:$2.0万
-
财政年份:2014
-
负责人:Michela Taufer
-
依托单位:
SHF: Small: Collaborative Research: Modeling and Analyzing Big Data on Peta- and Exascale Distributed Systems supported by MapReduce Methodologies
-
批准号:1318445
-
项目类别:Standard Grant
-
资助金额:$42.7万
-
财政年份:2013
-
负责人:Michela Taufer
-
依托单位:
CSR: Small: Collaborative: Pursuing High Performance on Clouds and Other Dynamically Heterogeneous Computing Platforms
-
批准号:1217812
-
项目类别:Standard Grant
-
资助金额:$18.45万
-
财政年份:2012
-
负责人:Michela Taufer
-
依托单位:
国内基金
海外基金
基于重要农地保护LESA(Land Evaluation and Site Assessment)体系思想的高标准基本农田建设研究
-
批准号:41340011
-
项目类别:专项基金项目
-
资助金额:20.0万元
-
批准年份:2013
-
负责人:钱凤魁
-
依托单位: