Developing complex scientific applications on High Performance Computing systems requires both domain knowledge and expertise in parallel and distributed programming models. Domain Specific Languages (DSLs) are a promising approach to hide the complexity of HPC systems, apply domain specific optimizations and boost programmer’s productivity.

The following figure shows how we think that applications should be designed:

Our vision is that applications should be written using DSLs, which have a high-level syntax and domain optimizations. Internally, these DSLs should be implemented using programming models that hide the complexity related to the hardware where these applications will be executed.