Evaluation and Deployment of ParaScope

ParaScope 的评估和部署

• 批准号: 9201498
• 负责人: Malvin Kalos
• 金额: --
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1992
• 资助国家: 美国
• 起止时间: 1992-09-01 至 1994-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9201498&HistoricalAwards=false
• 关键词: Evaluation; Deployment; ParaScope

The Center for Theory and Simulation in Science and Engineering at Cornell University (Cornell Theory Center) proposes a collaboration with the Center for Research on Parallel Computation (CRPC), headquartered at Rice University, for research in the development of scaleable parallel programs in a scientific programming environment. Ken Kennedy;s research group at Rice has developed a prototype parallel programming environment called ParaScope ?4,5,6!. Cornell is one of about twenty sites that have received this software for testing and evaluation. We propose to work together to evaluate and deploy this software. The goal of the proposed research is twofold: (1) to see how useful such an environmen is for scientific programming, and (2) to determine how extensible such and environment is, in scaleability and retargeting to different manchines. The time is ripe for such a study because of the tool's maturity and because of the growing need to retarget user programs quickly to new systems. The High Performance Computing and Communications Initiative will further encourage the proliferation of parallel systems as new approaches to parallel computing are explored. For the next five years, rapid change in our field demands parallel programming tools such as Parascope that can be quickly adapted to support new target parallel architectures.

康奈尔大学科学与工程理论与模拟中心（康奈尔理论中心）提议与总部位于莱斯大学的并行计算研究中心（CRPC）合作，研究在科学编程环境中开发可扩展的并行程序。 Rice 的 Ken Kennedy 研究小组开发了一个名为 ParaScope ?4,5,6! 的原型并行编程环境。 康奈尔大学是大约二十个已收到该软件进行测试和评估的站点之一。 我们建议共同评估和部署该软件。 拟议研究的目标有两个：（1）了解这样的环境对于科学编程有多大用处，（2）确定这样的环境在可扩展性和重新定位到不同机器方面的可扩展性。 由于工具的成熟以及将用户程序快速重新定位到新系统的需求不断增长，进行此类研究的时机已经成熟。 随着并行计算新方法的探索，高性能计算和通信计划将进一步鼓励并行系统的扩散。 未来五年，我们领域的快速变化需要并行编程工具，例如 Parascope，它可以快速适应新的目标并行架构。