The growing gap between I/O bandwidth and compute speed, and the need to significantly improve system resiliency are two of the most important Exascale challenges. DEEP-ER will extend the Cluster-Booster architecture of the DEEP project with a highly scalable I/O system and implement an efficient mechanism to recover application tasks that fail due to hardware errors. As a result, I/O parts of HPC codes will run faster and scale up better. These benefits will be demonstrated with the optimisation of seven HPC applications coming from different scientific and engineering areas and representing requirements of simulation-based and data-intensive HPC codes. The prototype developed and constructed by the DEEP-ER project will be used to prove the viability of the concept for 500 Petaflop class of supercomputers.