ProtoDUNE and DUNE computing will require a sizable and reliable hardware and software technology infrastructure. Nektarios Benekos of CERN is leading the effort for providing this infrastructure for the Neutrino Platform Experimental Hall, where the two ProtoDUNE detectors will reside.
“Experiments have always faced computing challenges, and their needs tend to outpace available and affordable solutions,” said Benekos. “The lifetimes of large science projects like DUNE are much longer than any given cycle in hardware and software technology. The ProtoDUNE experiments are preparing the S&C groundwork for DUNE so that it can meet these challenges effectively.”
Benekos developed and maintains a computer cluster of approximately 300 nodes with sufficient storage capacity to run the dedicated LArSoft and DUNETPC software for data analysis. He will install part of the cluster next March close to the experimental hall for ProtoDUNE-SP prompt processing needs. A dedicated fiber network infrastructure will be in place for a high transfer bandwidth of 40 Gbps and he expects to store the data in EOS-neutplatform space for further processing. Along with Maxim Potekhin from BNL, Benekos is currently polling various groups in DUNE to assess their needs.
“I want to to understand the physicists’ needs and ensure that a workable system gets put in place,” he said.