Acquisition of a High Performance Computing Cluster for the Geophysics Group at New Mexico State University

为新墨西哥州立大学地球物理小组收购高性能计算集群

• 批准号: 1661985
• 负责人: Lauren Waszek
• 金额: $ 7.5万
• 依托单位: New Mexico State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-08-15 至 2018-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1661985&HistoricalAwards=false
• 关键词: Acquisition; Performance; Computing; Cluster; Geophysics

This award will support the acquisition of a High Performance Computer (HPC) cluster for the Geophysics group at New Mexico State University (NMSU), and will benefit a range of geophysical, material, and solar research. NMSU is identified as a Hispanic Serving Institution, and approximately half of enrolled students are of Hispanic origin. The cluster will add new teaching opportunities in computational modeling, leading to better professional preparation of undergraduate and graduate students, as well as training the future generation of geoscientists. The work supported by this project will also facilitate redeveloping a new Geophysics curriculum. Research will model the structure, composition, and processes of Earth's surface and interior; highly complex problems requiring computationally intensive analysis. Planned projects include inversions to obtain models of Earth structures, calculations for synthetic data, and computational mineral physics calculations to explore the crystal structure of Earth and geo-inspired materials. The research is well aligned with the NSF mission to promote the progress of science via advancing our knowledge of our planet and the sun. New models of Earth's structure inform regarding potential earthquake hazard, and the novel ground motion modeling techniques employed here also benefit nuclear monitoring. Geo-inspired materials design will address technological solutions to societal challenges, with impact across engineering and material science disciplines. Further research that will benefit from the HPC includes the structure of the sun and other planets, providing numerous multidisciplinary opportunities.The seismic properties of Earth's interior are highly heterogeneous across multiple length scales. The structures must be constrained as accurately as possible in order to understand the underlying geodynamical processes and material composition. This necessitates various geophysical methodologies which require parallel computing, including waveform inversions for seismic velocity, anisotropy, and attenuation structure, forward modeling of full seismic waveforms using 3D Earth models, and calculations of earthquake and nuclear explosion ground motion. The resultant highly detailed seismic models will provide essential information for the next generation of geodynamical simulations, and mineral physics studies, as well as implications for earthquake hazard and nuclear monitoring. The massive quantities of seismic data utilized in our projects also entail highly computationally intensive processing methods, such as waveform correlations and modeling, stacking of real and synthetic array data, and large error calculations. In addition to modeling crystal structures within Earth, the computational mineral physics calculations will explore novel geo-inspired technological materials with applications to material science, engineering, and condensed matter physics. Solar physics research into interior plasma flows that transport magnetic flux through the star is crucial for understanding the solar dynamo. This new cluster facility will also benefit studies of the solar atmosphere, especially the solar flare prediction, which can drastically affect the communication satellites and electronics equipment on Earth.

该合同将支持新墨西哥州州立大学（NMSU）地球物理学小组购买高性能计算机（HPC）集群，并将使一系列地球物理、材料和太阳能研究受益。NMSU被确定为西班牙裔服务机构，大约一半的入学学生是西班牙裔。该集群将在计算建模方面增加新的教学机会，从而为本科生和研究生提供更好的专业准备，并培训未来一代的地球科学家。该项目所支持的工作还将促进重新编制新的地球物理学课程。研究将模拟地球表面和内部的结构，组成和过程;需要计算密集型分析的高度复杂的问题。计划中的项目包括反演以获得地球结构模型，计算合成数据，以及计算矿物物理计算以探索地球和地质启发材料的晶体结构。这项研究与NSF的使命是一致的，即通过提高我们对地球和太阳的认识来促进科学的进步。地球结构的新模型告知潜在的地震危险，这里采用的新地面运动建模技术也有利于核监测。地理启发材料设计将解决社会挑战的技术解决方案，对工程和材料科学学科产生影响。从HPC中受益的进一步研究包括太阳和其他行星的结构，提供了许多多学科的机会。地球内部的地震特性在多个长度尺度上是高度异质的。必须尽可能精确地约束结构，以便了解潜在的地球动力学过程和物质组成。这需要各种地球物理方法，需要并行计算，包括地震速度，各向异性和衰减结构的波形反演，使用三维地球模型的全地震波形的正演模拟，以及地震和核爆炸地面运动的计算。由此产生的非常详细的地震模型将为下一代地球动力学模拟和矿物物理学研究提供必要的信息，并对地震灾害和核监测产生影响。在我们的项目中使用的大量地震数据还需要高度计算密集型的处理方法，例如波形相关和建模、真实的和合成阵列数据的叠加以及大误差计算。除了对地球内部的晶体结构进行建模外，计算矿物物理学计算还将探索新的地质启发技术材料，并将其应用于材料科学，工程和凝聚态物理学。对在星星中传输磁通量的内部等离子体流的太阳物理学研究对于理解太阳发电机至关重要。这一新的集群设施还将有利于对太阳大气层的研究，特别是太阳耀斑的预测，因为太阳耀斑可能对地球上的通信卫星和电子设备产生巨大影响。