Deadlock-Free Fault Tolerant Wormhole Routing in Mesh based Massively Parallel Systems
Johan Vounckx, Geert Deconinck, Rudy Lauwereins, Jean Peperstraete · 1994
Abstract: In this paper we present a routing scheme which is extremely suited for use in massively parallel systems. The routing algorithm is fault-tolerant so that network failures will not stop the system. For reasons of scalability, the routing information is extremely compact, also when the network is injured. The wormhole routing technique guarantees very low routing latency. Moreover the routing is guaranteed to be deadlock-free without the need for dedicated buffering. Key words: compact routing, wormhole rousting, deadlock-free routing, fault tolerance, massively parallel systems Nowadays challenge in supercomputing Is to reach the teraflop goal to deal with the grand challenge applications. Despite the ever increasing performance of the processors, massively parallel systems must still be used. Thousands of processors must work in parallel to obtain this enormous computational power. Not only more processors should be used, the performance of the communication network should also scale. Small delays are