Extracting Characteristics from Functional Programs for Mapping to Massively Parallel Machines

For problems with highly dynamic behavior, our experiments showed that there are speciic characteristics for diierent applications. We therefore propose a mapping environment providing several strategies for both granularity control and dynamic load balancing. For appropriate selection and parameterization of strategies, we extract the application characteristics using trace-based prooling and appropriate evaluations. Our prooling approach diiers from others in that more information (like argument sizes or the branching factor) is collected and more advanced evaluations are performed. Large real-life applications have been successfully measured, and examples demonstrate diierences in characteristics and which the corresponding strategies being appropriate are.