DEEP-ER

The goal of DEEP-ER is to update the Cluster-Booster architecture introduced by the DEEP project and extend it with additional parallel I/O and resiliency capabilities. In DEEP-ER, Cluster and Booster nodes will be connected to a uniform network. Novel non-volatile memory technology will be tested at different architecture levels, building up a multi-level storage hierarchy. A software environment for parallel I/O and resiliency will be built upon this storage infrastructure. To recover applications after hardware failures, a multi-level checkpoint/restart mechanism using the available memory devices and a task-based recovery mechanism based on the OmpSs programming environment will be set up.