Evaluating Memory System Performance of a Large-Scale NUMA Multiprocessor

The e ectiveness of large scale computing depends to a great extent on the performance of the memory system. As shared memorymultiprocessors grow in size, their memory hierarchy deepens, resulting in a design with non-uniform latencies. In this paper, we explore the implications of multi-valued memory latencies. In particular, we study the e ect of a nonuniform tra c distribution on a hierarchical large scale NUMAmultiprocessor named Hector. Memory analysis is of interest because memory is a frequent source of poor performance in large scale multiprocessors. We have developed an analytical model that includes the e ects of increased contention for system resources, and the impact of the arbitration algorithm on the network tra c. Our analysis has been validated with a detailed simulator. Also, we have examined two techniques for reducing memory latency. We assess the potential performance gains from replication of data and investigate the improvement in memory utilization by allowing memory request bu ering. Furthermore, we studied the sensitivity of the memory performance to changes in background tra c. We found that inter-station tra c has a signi cant performance e ect.