The Effects of Network Contention on Processor Allocation Strategies

Various processor allocation strategies have been proposed for scalable parallel computers (SPCs). These strategies try to maximize the overall system utilization and, in the mean time, try to avoid network contention among different processor partitions. This paper provides an intensive simulation study investigating whether contention-free processor allocation strategies are indeed important. Our simulation considers both meshand MIN-based wormhole parallel computers, the communication characteristics of individual applications, and the impact due to communication software latency. We show that for systems with high software latency, there is no need of contention-free processor allocation policies. However, if the software latency is very small or the message size is very long, contention-free allocation policies should be developed.