Review of real-time data processing for collider experiments

Abstract

We review the status of, and prospects for, real-time data processing for collider experiments in experimental High Energy Physics. We discuss the historical evolution of data rates and volumes in the field and place them in the context of data in other scientific domains and commercial applications. We review the requirements for real-time processing of these data, and the constraints they impose on the computing architectures used for such processing. We describe the evolution of real-time processing over the past decades with a particular focus on the Large Hadron Collider experiments and their planned upgrades over the next decade. We then discuss how the scientific trends in the field and commercial trends in computing architectures may influence real-time processing over the coming decades.