|
      
|
|
|
|
Algorithmic skeletons abstract commonly used patterns of parallel computation, communication, and interaction. They present a top-down transformational approach where programs are formed from skeleton nesting. By demonstrating a predictable communication and computation structure, they provide a preponderant foundation for performance modelling and estimation.
Scant research has been conducted on using this predicting information to enhance performance in heterogeneous environments. We propose the use of these predicting properties to adaptively enhance the performance of skeletons, in particular of a processor farm, within a computational grid.
We present some preliminary results of parallel executions in a grid. Our approach provides an adaptive systems-infrastructure model, using the Network Weather Service to provide bandwidth, latency and processor measurements. The implementation is based on C and MPI.
