Evolution of the CMS Submission Infrastructure to support heterogeneous resources in the LHC Run 3

The landscape of computing power available for the CMS experiment is rapidly evolving, from a scenario dominated by x86 processors deployed at WLCG sites towards a more diverse mixture of Grid, HPC, and Cloud facilities, incorporating a higher fraction of non-CPU components such as GPUs. Using these facilities heterogeneous resources efficiently to process the vast amounts of data to be collected in the LHC Run3 and beyond, is key to CMS achieving its scientific goals. The Submission Infrastructure is the main component of the resource acquisition and workload to resource matchmaking systems in CMS Offline Computing. It is therefore firstly mandatory to adapt this infrastructure to be able to request and aggregate heterogeneous resources from our providers and to integrate them into the CMS HTCondor infrastructure. Secondly, it is crucial to optimize the matchmaking of CMS workloads to heterogeneous resources, in order for CMS to succeed in effectively exploiting the enormous amount of computing power that is expected to be available in the form of GPUs. This report presents how this has been technically achieved, as well as a brief description of the already existing pool of GPUs ready for CMS use.