课题基金 / 基金详情

Supporting and Enhancing the HPC Challenge Benchmark for Hybrid-Multicore Computers

Supporting and Enhancing the HPC Challenge Benchmark for Hybrid-Multicore Computers
支持和增强混合多核计算机的 HPC 挑战基准
批准号:
1038814
负责人:
Jack Dongarra
金额:
$40.0万
依托单位国家:
美国
项目类别:
Standard Grant
财政年份:
2011
资助国家:
美国
项目状态:
已结题
起止时间:
2011-03-01 至 2015-02-28

项目摘要

项目成果

Jack Dongarra的其他基金

相似基金

相关文献

中文摘要
翻译
点击翻译按钮获取中文摘要
英文摘要
The tight coupling that now exists between accelerating progress in nearly every critical scientific field and increasing the availability of high performance computing (HPC) means that agencies committed to supporting leading edge research will have to continue, or even expand, their investment in high end computing capacity. Unfortunately, the radical changes in system architecture that new extreme scale platforms are bringing with them (massive concurrency, processor heterogeneity, constrained power budgets, and complex memory architectures) are also crippling our ability to make such investment decisions in a rational and well-informed way. In particular, tools such as the HPC Challenge (HPCC) benchmark suite, which have played a vitally important role in evaluating the impact of system design choices on the performance of scientific applications, are not as well adapted as they need to be to the incredible complexities of emerging supercomputing platforms. Our proposed plan for supporting and enhancing the HPCC benchmark for hybrid-multicore computers is designed to maintain and extend the effectiveness of this essential tool.Accordingly, our plan for updating and enhancing HPCC will focus on three main areas of activity: 1) investigate new rules for benchmark execution; 2) improve scalability of benchmark for future architectures; and 3) expand community interaction.In terms of contributing to national cyberinfrastructure, the planned work builds upon and strengthens our ongoing collaborations with NSF Centers and HPC vendors. We are involved in the NSF Centers and have a wide collection of users and applications to draw on. These facilities contribute software and expertise to HPCC and will provide an early operational testbed for new technologies and application concepts to be developed for HPCC. Responses to our plans from vendors have been encouraging. With their cooperation, we will produce software that will be transitioned to the commercial software vendors to complete the development that we have initiated and provide these new capabilities to their customers in the research community.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
Travel: Workshop on Clusters, Clouds, and Data Analytics for Scientific Computing 2024
  • 批准号:
    2336813
  • 项目类别:
    Standard Grant
  • 资助金额:
    $2.5万
  • 财政年份:
    2023
  • 负责人:
    Jack Dongarra
  • 依托单位:
Workshop on Clusters, Clouds, and Data Analytics for Scientific Computing
  • 批准号:
    2001329
  • 项目类别:
    Standard Grant
  • 资助金额:
    $2.0万
  • 财政年份:
    2020
  • 负责人:
    Jack Dongarra
  • 依托单位:
Workshop on Clusters, Clouds, and Data Analytics in Scientific Computing
  • 批准号:
    1800946
  • 项目类别:
    Standard Grant
  • 资助金额:
    $1.93万
  • 财政年份:
    2018
  • 负责人:
    Jack Dongarra
  • 依托单位:
Toward a common digital continuum platform for big data and extreme-scale computing (BDEC2)
  • 批准号:
    1849625
  • 项目类别:
    Standard Grant
  • 资助金额:
    $20.34万
  • 财政年份:
    2018
  • 负责人:
    Jack Dongarra
  • 依托单位:
海外基金