HEP computing collaborations for the challenges of the next decade

Large High Energy Physics (HEP) experiments adopted a distributed computing model more than a decade ago. WLCG, the global computing infrastructure for LHC, in partnership with the US Open Science Grid, has achieved data management at the many-hundred-Petabyte scale, and provides access to the entire community in a manner that is largely transparent to the end users. The main computing challenge of the next decade for the LHC experiments is presented by the HL-LHC program. Other large HEP experiments, such as DUNE and Belle II, have large-scale computing needs and afford opportunities for collaboration on the same timescale. Many of the computing facilities supporting HEP experiments are shared and face common challenges, and the same is true for software libraries and services. The LHC experiments and their WLCG- partners, DUNE and Belle II, are now collaborating to evolve the computing infrastructure and services for their future needs, facilitated by the WLCG organization, OSG, the HEP Software Foundation and development projects such as HEP-CCE, IRIS-HEP and SWIFT-HEP. In this paper we outline the strategy by which the international HEP computing infrastructure, software and services should evolve through the collaboration of large and smaller scale HEP experiments, while respecting the specific needs of each community. We also highlight how the same infrastructure would be a benefit for other sciences, sharing similar needs with HEP. This proposal is in line with the OSG/WLCG strategy for addressing computing for HL-LHC and is aligned with European and other international strategies in computing for large scale science. The European Strategy for Particle Physics in 2020 agreed to the principles laid out above, in its final report.