![]() ![]() Detailed simulation results show that the hyperswitch network is cons&entiy more efficient than fized path routing for large message traffic conditions. ![]() This method can be accomplished in a static topology such as the hypercube network if the nodea have switching elements which are capable of performing the necessary rouh'ng header retions dynamically. Here, available or fault free paths need not be specified by a source because the routing header can be modified in response to congestion or faults encountered a9 a path is established. The method presented in this paper realizes a kind of interconnectibn network called a hyperswitch network, that is achieved using a m&we of static and dynamic topologies. 91109 University of Illinois, Urbana-Champaign ABSTRACT The performance of a parallel algorithm depends in a large part on the interconnection topology of the multicomputer system. Peterson Jet Propulsion Laboratory, California Institute of Technology Pasadena, Calif. INTRODUCTION A multicomputer interconnection network may The simulation results also show that the hyperswitch network haa ep'valent messages with localized and latency overhead for antilocal destinations (i.e., less then a 25% difference between diameter i and 5). Hyperswitch network for the hypercube computer Hyperswitch network for the hypercube computerĬhow, E. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |