Speaker
Description
The Deep Underground Neutrino Experiment (DUNE) will deploy four 10 kt
fiducial mass time projection chambers in order to study accelerator neutrinos,supernova neutrinos, beyond the standard model physics, atmospheric neutrinos,and solar neutrinos. Reconstructing data in varying time domains over the nearly 400,000 channels needed to monitor this volume presents the complex challenge of processing this information in flexible data-grain sizes suitable for affordable compute nodes. Phlex, a framework under development, will address this challenge using new I/O and storage infrastructure called the Fine-grained Object Reading/writing Model (FORM). In addition to providing flexible, grained I/O, FORM will support several storage technologies, including ROOT’s legacy TTree and its replacement RNTuple. This poster will show first results for RNTuple storage for data of varying grain sizes.