M. Essaïdi, I. Guérin Lassous, and J. Gustedt (France)
Coarse grained parallel models, parallel architectures, communication interfaces, complexity analysis.
We present SSCRAP, an environment for the development of coarse grained algorithms. SSCRAP is well suited for regular as well as irregular problems and implements the currently known coarse grained models for distributed and shared memory architectures. It allows the implementation of algorithms designed on a high level of abstraction and –at the same time– handles communication and synchronization aspects efficiently. By making the communication level transparent for the user it guarantees a good portability. SS CRAP also provides a detailed execution analysis: users are able to compare efficiency of algorithms and to check the predictability level of the considered theoretical complexity.
Important Links:
Go Back
