The computing of the Lhc experiments

The Lhc experiments have thousands of collaborators distributed worldwide who expect to run their physics analyses on the collected data. Each of the four experiments will run hundreds of thousands of jobs per day, including event reconstruction from raw data, analysis on skimmed data, and production of simulated events. At the same time tens of peta bytes of data will have to be easily available on a complex distributed computing fabric for a period of at least ten years. These challenging goals have prompted the development and deployment of reliable Grid services, which have been thouroughly tested and put at the needed scale over the last years. This paper concentrates on CMS computing needs for data taking at Lhc and highlights the INFN-Grid contribution to the effort.