原子探针层析成像数据分析的强标度和开源工具
On Strong Scaling and Open Source Tools for Analyzing Atom Probe Tomography Data
摘要:
原子探针层析成像(APT)已经发展成为一种多功能的纳米分析表征工具,其应用范围从材料科学到地质学甚至更广。全世界已经有超过100台APT显微镜。来自APT数据的信息需要对重建的点云进行后处理,这是通过数据科学方法的基本实现来实现的,主要是用专有软件来执行。软件的局限性促使APT社区开发补充后处理工具,以应对日益增加的方法复杂性和更高的质量要求:例如如何提高方法透明度,如何支持批处理能力,以及如何更完整地记录方法和计算工作流程,以更好地符合FAIR数据管理原则。APT软件工具领域的一个差距是支持科学计算硬件的开放工具的集合。在这里,我们介绍PARAPROBE,一个开源的,有效的工具,用于科学计算APT数据。我们展示了如何处理几个计算几何,空间统计和聚类任务性能的数据集高达20亿离子。我们的并行化工作产生了数量级的性能增益,并提供了批处理能力。我们贡献这些工具,以努力开放APT数据挖掘,并简化它,使严格的量化,敏感性分析和跨方法基准测试的工具提供给从业者。
英文摘要:
Atom probe tomography (APT) has matured to a versatile nanoanalytical characterization tool with applications that range from materials science to geology and possibly beyond. Already, well over 100 APT microscopes exist worldwide. Information from the APT data requires a post-processing of the reconstructed point cloud which is realized via basic implementations of data science methods, mostly executed with proprietary software. Limitations of the software have motivated the APT community to develop supplementary post-processing tools to cope with increasing method complexity and higher quality demands: examples are how to improve method transparency, how to support batch processing capabilities, and how to document more completely the methods and computational workflows to better align with the FAIR data stewardship principles. One gap in the APT software tool landscape has been a collection of open tools which support scientific computing hardware. Here, we introduce PARAPROBE, an open source, efficient tool for the scientific computing of APT data. We show how to process several computational geometry, spatial statistics, and clustering tasks performantly for datasets as large as two billion ions. Our parallelization efforts yield orders of magnitude performance gains and deliver batch processing capabilities. We contribute these tools in an effort to open up APT data mining and simplify it to make tools for rigorous quantification, sensitivity analyses, and cross-method benchmarking available to practitioners.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。