Next Generation Parallel Computing Infrastructure for Numerical Relativity

下一代数值相对论并行计算基础设施

• 批准号: 0929114
• 负责人: Yosef Zlochower
• 金额: $ 7万
• 依托单位: Rochester Institute of Tech
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-09-01 至 2011-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0929114&HistoricalAwards=false
• 关键词: Next; Generation; Parallel; Computing; Infrastructure

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). This award supports research to develop software infrastructure required to efficiently solve the Einstein Equations of General Relativity on the next generation of supercomputers. This capability will then be exploited to evolve systems containing two black holes of unequal mass, an important potential source of gravitational waves detectable by ground-based and space-based detectors such as LIGO and LISA. Computer simulations of binary black holes provide gravitational waveforms that can be used to search for weaker signals than could be found without this information. The new supercomputers achieve speed breakthroughs by increasing the number of cores per CPU to tens or even hundreds. Computer codes designed for today's computers must be redesigned to take maximum advantage of the new architechture. This project is part of an ongoing effort at RIT to develop new, highly efficient, high-accuracy numerical algorithms for General Relativity.

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。该奖项支持研究开发在下一代超级计算机上有效求解广义相对论爱因斯坦方程所需的软件基础设施。然后，这种能力将被利用来发展包含两个质量不等的黑洞的系统，这是LIGO和丽莎等地基和天基探测器可探测到的引力波的重要潜在来源。双黑洞的计算机模拟提供了引力波形，可以用来搜索比没有这些信息时更弱的信号。新的超级计算机通过将每个CPU的核心数量增加到数十个甚至数百个来实现速度突破。为今天的计算机设计的计算机代码必须重新设计，以最大限度地利用新的体系结构。这个项目是RIT正在进行的努力的一部分，为广义相对论开发新的，高效的，高精度的数值算法。