IPDPS '05 Proceedings of the 19th IEEE International Parallel and Distributed Processing Symposium (IPDPS'05) - Workshop 14 - Volume 15
IPDPS '05 Proceedings of the 19th IEEE International Parallel and Distributed Processing Symposium (IPDPS'05) - Workshop 14 - Volume 15
| Hi-index | 0.00 |
The Multiple Associative Computing (MASC) model is a generalized associative-style parallel computation model that was developed at Kent State University in 1990s. It provides a complete paradigm that can be used for both special-purpose and general-purpose parallel computation. The primary focus in this research is to evaluate the power of the MASC model and to provide a better understanding of it. This work consists of three parts. The first part is to justify the timing assumptions for the basic associative operations of the MASC model in order to establish a firm theoretical foundation for the MASC model. The associative operations are used extensively in MASC algorithms. The second part creates simulations between the MASC model and well-known parallel enhanced mesh models. The simulations of the enhanced mesh models provide an efficient method for converting algorithms designed for enhanced meshes to the MASC model. The third part involves using the ASC model (i.e., MASC with one instruction stream or an associative SIMD computer) to provide an efficient polynomial time solution to the real-time Air Traffic Control problem. However, this type of real-time problem is currently considered to be unsolvable in polynomial time using a MIMD computer (i.e. a multiprocessor). Since the MIMD model of parallel computers is generally believed to be more powerful than the SIMD model due to the fact of MIMD's asynchronous execution, the preceding anomaly indicates that this belief needs to be reconsidered.