Towards a shared-memory massively parallel multiprocessor

Abstract

A set of ultra high throughput (more than one Gigabits per second) serial links used as processor-memory network can lead to the starting up of a shared-memory massively parallel multiprocessor. The bandwidth of the network is far beyond values found in present shared-memory multiprocessor networks. To feed this network the memory must be serially multiported. Such a multiprocessor can actually be build with current technologies.This paper analyzes the characteristics of such a novel architecture, presents the solutions that must be considered and the practical problems associated with close of experiments.These results show then the way to effectively build this multiprocessor, taking into account main topics such as data coherency, latency time and scalability.