On Strong Scaling Open Source Tools for Mining Atom Probe Tomography Data

There is an increasing demand in the atom probe tomography (APT) community for accurate and efficient open source tools for assessing the quality and content of increasingly large experimental datasets, combined with a mindset change towards open source reproducible methods [1]. APT is a microscopy and microanalysis technique which produces datasets in the form of point clouds representing a material’s volume at the nanoscale with positional and chemical information for each ion. This information allows to characterize microstructures and portray snapshots of their evolutionary mechanisms. Currently, two software tool options exists: one is IVAS [2], the Integrated Visualization and Analysis Software, a functionally rich graphical front end solution; the other, complementary option, are scripts that are less functionally diverse, but cutting-edge proof-of-concept tools, developed by experimentalists [3,4] and primarily, like IVAS, using sequential algorithms and implementations. IVAS is proprietary, not cross-compatibly executable, and closed source which poses an effective barrier to rigorous functional verification and extension. In contrast, the source code to these individual, ad-hoc tools is usually open, they employ fully transparent algorithms and thus patch and complement functionalities that are missing in IVAS.