Using nodes in a cluster efficiently
The purpose of this paper is to evaluate how to use nodes in a cluster efficiently by studying the NAS Parallel Benchmarks (NPB) on Intel Xeon and AMD Opteron dual CPU Linux clusters. The performance results of NPB are presented both with one MPI process per node (1 ppn) and with two MPI processes per node (2 ppn). One would like to run all applications on a cluster with two processors per node using 2 ppn instead of 1 ppn in order to utilize the second processor on each node. However, the performance results from running the NPB and from the memory bandwidth benchmarks show that better performance can sometimes be achieved using 1 ppn. Our performance results show that the Opteron/Myrinet cluster is able to achieve significantly better utilization of the second processor than the Xeon/Myrinet cluster.