Exploitation of the MareNostrum 4 HPC using ARC-CE

Abstract

The resources of the HPC centers are a potential aid to meet the future challenges of HL-LHC [1] in terms of computational requirements. Spanish HPC centers have recently been used to implement all necessary edge services to integrate resources into the LHC experiment workflow management system. In this article, we describe the integration of ATLAS with the extension plan to other LHC experiments. We chose to configure a dedicated ARC-CE [2] and interact with the HPC login and transfer nodes using ssh commands. The repository that includes a partial copy of the ATLAS experiment software on CVMFS is packaged in a singularity image to overcome network isolation for HPC nodes and reduce software requirements. ATLAS provided the initial container, and the authors adapted it to the specific HPC environment. This article shows the Spanish contribution to the simulation of experiments after the Spanish Ministry of Science agreement and the Barcelona Supercomputing Center (BSC), the center that operates MareNostrum 4. Finally, we discuss some challenges to take advantage of the next generation of HPC machines with heterogeneous architecture combining CPU and GPU.