Simplifying Software Development and Increasing Software Productivity on High End Computers

The complexity of modern scientific software is an important issue that needs to be addressed as part of an effort to revitalize high end computing (HEC). The complexity arises from both the scientific demands and the increasingly complex computers on which the software is run. We believe that HEC software developers need to be able to use a higher level of abstraction to express their computational problems. New programming models and technologies such as domain-specific languages coupled with automatic code generation, source-to-source translation tools, and component models are all sufficiently advanced that a concerted investment aimed at bringing them to HEC computational scientists will bring significant near-term payoffs.