Revisiting communication performance models for computational clusters

—In this paper, we analyze restrictions of traditional models affecting the accuracy of analytical prediction of the execution time of collective communication operations. In particular, we show that the constant and variable contributions of processors and network are not fully separated in these models. Full separation of the contributions that have different nature and arise from different sources will lead to more intuitive and accurate models, but the parameters of such models cannot be estimated from only the point-topoint experiments, which are usually used for traditional models. We are making the point that all the traditional models are designed so that their parameters can be estimated from a set of point-to-point communication experiments. In this paper, we demonstrate that the more intuitive models allow for much more accurate analytical prediction of the execution time of collective communication operations on both homogeneous and heterogeneous clusters. We present in d...

• Type: Conference
• Year: 2009
• Where: IPPS
• Authors: Alexey L. Lastovetsky, Vladimir Rychkov, Maureen O'Flynn
• Keywords: Collective Communication Operations, Distributed And Parallel Computing, Execution Time, IPPS 2009, Traditional Models