Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) Final Report

doi:10.2172/1136155

GSTDTAP > 地球科学

DOI	10.2172/1136155
报告编号	LLNL-TR-655463
来源ID	OSTI ID: 1136155
	Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) Final Report
	Williams, Dean N.
	2014-05-19
出版年	2014
页数	12
语种	英语
国家	美国
领域	地球科学
英文摘要	A partnership across government, academic, and private sectors has created a novel system that enables climate researchers to solve current and emerging data analysis and visualization challenges. The Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) software project utilizes the Python application programming interface (API) combined with C/C++/Fortran implementations for performance-critical software that offers the best compromise between "scalability" and âease-of-use.â The UV-CDAT system is highly extensible and customizable for high-performance interactive and batch visualization and analysis for climate science and other disciplines of geosciences. For complex, climate data-intensive computing, UV-CDATâs inclusive framework supports Message Passing Interface (MPI) parallelism as well as taskfarming and other forms of parallelism. More specifically, the UV-CDAT framework supports the execution of Python scripts running in parallel using the MPI executable commands and leverages Department of Energy (DOE)-funded general-purpose, scalable parallel visualization tools such as ParaView and VisIt. This is the first system to be successfully designed in this way and with these features. The climate community leverages these tools and others, in support of a parallel client-server paradigm, allowing extreme-scale, server-side computing for maximum possible speed-up.
URL	查看原文
来源平台	US Department of Energy (DOE)
引用统计
文献类型	科技报告
条目标识符	http://119.78.100.173/C666/handle/2XK7JSWQ/6747
专题	地球科学
推荐引用方式 GB/T 7714	Williams, Dean N.. Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) Final Report,2014.